I would point out that the passage does not say at all that the beneficial side-effect can be reproduced (or at least, reproduced without further experimentation). Could you give other references supporting your point more explicitly?
Once again, keep in mind that the core itself explicitly says that "learning" spells from a source is more of a process of "reinventing with guidance" than "being taught".
Um... sorry, E - that's not what it says at all.
First, if you are "taught" a spell it is identical, except for the wiz's sigil (p 95, col i), so that's one source that isn't covered.
But more, that's not what the Core Rulebook implies, much less "explicitly says". In fact, it says exactly the opposite, that
"This (effect) is (except for sigil) an exact reproduction of the original effect" [list][list][list][list][list][list][list][list][list](p 101, col 1, par ii, emphasis added).[/list:u][/list:u][/list:u][/list:u][/list:u][/list:u][/list:u][/list:u][/list:u]
That's something that has never been made perfectly clear.
I remember in an older edition of the rules (3rd?), that table suggested to "increase the effect by a magnitude (or two?)" (or words to that effect) - strongly implying (without actually stating) that a 20th level spell could produce a 25th level effect.
Ysmv, but I find this uncomfortable on several levels.
Are the side effects large and/or strong enough to warrant a greater magnitude or additional Art that is just not there in the final math?
If it was just "thrown in for free" and then anyone else could learn it, there should be examples of that in canon, or at the very least some hint or footnote somewhere that suggests it's possible - and subject to the above clarification, there simply are not.
If the "side effect" is noticeably stronger, then the spell should be built around the total effect, not have the final effect ignored and shoe-horned into the original TeFo magnitude. (And, yes, with unexpected Arts added as requisites/etc., this could mean that a spell is invented that the Experimenter could not have invented, and perhaps cannot even cast reliably. Such is experimentation.)
Another solution would be to rule that, if a Lab Text has been created via Experimentation, it can likewise only be "learned" with experimentation. Might be better, might be worse, but that takes the "identical" problem out of the equation. (However, that does not solve the "Teaching a Spell" scenario.)
Well, I suppose that the actual "it' shows up in canon" is the one already mentioned - the Bonisagus' "cool spell research" book.
But yeah - the main issue I've had with that is that it doesn't show up any other place. And you'd think, with a new version of that book circulating every seven years, at least THOSE experimental spells would show up in more wizard's spell lists.
Now - from a meta standpoint, I can take a reasonable guess - it's an additional layer of complication that the writers didn't want to take for character creation - as it would require there be a "common core" of experimental spells (such as, say, from the aforementioned Bonisagus book) that folks learn from; and then the writers would have to determine which versions hold what spells, if the magi they're creating would have access to it, so on and so forth. And then they'd have to determine which versions have been superseded by advances in Magic Theory Integrations, etc. So probably it isn't worth the trouble.
THAT BEING SAID - it'd be a nice web-research project to through and write up, say the last 100 year's worth of Colentes Arcanorum, with just that information in it.
(As a simplification, you could assume that there is a lessening of interest the further you go back - as Breakthrough Research spells no longer apply, as their Breakthrough has already been integrated.)
EDIT - personally, I just HR that a magus can know a couple of experimental spells that their paren taught them - subject to reasonable storyguide limitations, each paren probably knows one or two that they'd be willing to teach. Beyond that, it's required Storyteller permission and whatnot.
EDIT II - I seem to recall contributing to at least one thread that was doing this last year... (here we go: https://forum.atlas-games.com/t/lets-design-a-folio-updated/8300/1
Let me quote from the exact passage you are referring to:
I think this is ultimately the crux of the problem.
If the bonus effect is "free" (say, your experimental MuCo(An) spell to turn into a wolf also allows you to talk to other animals without an Intellego requisite) and at the same time "transmissible" you run into problems of a) verisimilitude (why aren't the most commonly used spells "optimized"?) and b) complexity (what if people keep experimenting on experimental spells to pile bonus after bonus on them - say a spell to turn you into a wolf that also allows you to speak with animals and paralyze with your gaze and summon other wolves with your howl and command them and ...).
If the bonus is "free" but not "transmissible" it's a nice perk of experimentation for the experimenter, but causes no problems - it leaves the rest of the Hermetic landscape totally unaltered. This is what our troupe adopts, since we feel experimentation needs some "encouragement" - as it stands it's almost never worth it.
Alternatively, if the bonus is not free, then the question of whether it's transmissible becomes moot. You have gained a perfectly normal Hermetic spell (that others can learn as such); you just have learnt it in an unconventional way, possibly allowing you to learn something that you could not have otherwise learnt ... and that you may not even be able to cast. But again, except for the fact that you know it when you shouldn't, the Hermetic landscape is totally unaltered, so it's another "safe" way to play it.
But the fact that lab texts on "experimental" spells are circulated does not mean that, for the purpose of learning those spells, those lab texts are in any way different from "normal" lab texts. They could simply be circulated because they have value as sources of Insight in new directions that Magic Theory could take, or as reports of what one's colleagues managed to accomplish.
Also, it doesn't explain the nature of the experimentation. Does it apply to normal lab experimentation to push the individual[1] magus's boundaries or is it experimentation related to Original Research, and is the final product of research on a topic granting so many points towards a breakthrough. I tend to lean in towards the Original Research direction, myself. I don't recall why I lean this way, I would need to review the Original Research section to see if that's an accurate and reasonable understanding, if memory serves it has to do with how the final product of research can be taught, so it's reasonable to assume that the preceding steps are also teachable.
[1]I find this snippet from the main rule book enlightening.
That suggests that it is your magic, and your magic alone that allows you to do these things, and any lab text that is produced while it might describe the secondary effect simply won't work for anyone else the same way.
I always assumed the same. It's also a rationalization why many of the basic spells sometimes don't adhere to the given guidelines.
In that case, the author would have written "Original Research" or "Integration", rather than "experimental spell". Which isn't what is written. If you want to House Rule that, that's fine - but that's not what the text says. I agree that it would be more consistent if it were Integration or Original Research (which, if I recall the rules, explicitly can be taught) - but again: that's not what it says.
And if what you describe is the case, there would be no reason to put it in the book. "Here's a pillum of flame that the creator claims allows him to shoot three at a time! But you can't actually see that in the lab notes, so it's really just a pillum of flame." If what you said was the case, there should be some sort of bonus to experimental rolls based off of the lab notes - but there isn't.
Hi,
The side effect should:
Reflect the magus, sigil being a good place to start. Already mentioned.
Preferably be related to the Form of the spell.
Be balanced! Remember that every casting of this spell benefits from the side effect. Also, a major benefit is not a breakthrough.
Be fun.
Be apt. 
Even if the inventor is not Criamon. In a way, this sums it up. So there can be contradictions among the above.
Examples: Most magic towers are poor works of art, but this one is not. Studying a tower created by this spell for a season provides 10xp in Finesse; only one season per magus. The spell serves as a single good source of Insight for a breakthrough involving Hermetic Architecture, or something like that; it even acts as a tether that can increase Aura... but only at the first site, so much more research is needed. The tower is so perfect a defense that all attacks on it or its contents suffer one extra botch die and are stress (this is very, very good yet will not change the game and probably rarely comes into effect.)
Anyway,
Ken
As the original spell is CrTe, I would suggest:
The tower has huge windows of perfectly clear glass. Normal glass creation requires ludicrously high finesse rolls to make clear, so this is an impressive-looking effect.
The tower can also have a separate curtain wall as a bonus - normally creating a tower and a wall would be two spells or a group target, so this is a huge bonus but is flavourful. Alternatively, extra outbuildings for free.
+1
And to the other points, but this especially.
Tower as Tractatus - I LIKE it!
I didn't really make this up, though I'm happy to take credit. In A&A, studying a work of art can provide xp in the appropriate Ability. So a work of art created by Hermetic Magic ought to do the same, and the appropriate Ability is Finesse!
grin It also makes those stereotypical Jerbiton who spend seasons mooning over beautiful works of magic a little more dangerous and practical than they might seem....
I'd think it was cool if that was an option: say, if you open-end on the "Beneficial Side Effect" portion of the Experimental table - but ultimately it would probably throw the balance of Breakthrough research and Insight Integration off. Or if alternately it COULD be an Insight - but only a few points, and only in what amounts to a completely random Breakthrough that's potentially unrelated to your spell. Thus, someone would have to comb through all of the Experimental spells and find those that had similar breakthroughs, and combine them all into a greater whole.
That couldn't be hard, because they seem... 
For simplicity's sake, I agree with Ezzelino that having side benefits/flaws personnal to the magus who invented the spell should be favored. It also seems at to at least partly coincide with the rules.
Still, there's that bit about the collens. But...
Yeah, I said something like this elsewhere.
I like the idea that experimental spells sometime do "new things" that showcase that something you didn't know how to do, or didn't thought possible, are in fact doable with hermetic magic, providing insight for research in that direction.
So, I'd rule that, when experimentation yields a benefit, it may (list non exhaustive):
So, to come back to the topic, and try to follow my own advice:
Well, the way I see it, there is no balance to breakthroughs. As a game, AM is precariously balanced as it is, based on 5 editions of rl experience with it, and some might say that it is already imbalanced. Hermetic Breakthroughs make magic even more powerful, and in specific ways rather than across the board. Suddenly, X becomes "overpowered" and Y becomes worthless. Many breakthroughs are really campaign capstones or saga turning points.
Yeah I'm with those that would say Ars Magica is imbalanced and was never meant to be anything but. There are too many best choices, loopholes, and exploits in even the base magic system to claim Hermetic Magic is balanced. (Formulaic Army Killers, Pink Dot's, Creo Continent spells, Magic to Wealth Scheme's,Effect Expiry, Johnny One-Spells,FastCast/Multicast, Teaching/Writing Circles...) Since these are part of the basic magic system we tend to assume they are each dealt with somehow in the base setting. Either with story features or a "rule one magic don't work that way, sorry."
Since the "rules" of Magic are so closely tied to the setting, breakthroughs are quite literally game changers. Thus for me part of the point is to play out the setting coming to terms with a change in how magic woks. I like this sort of play but realize it does not have a place in all games not even all of my games.
Though I suppose there are plenty of potential Breakthroughs that would have almost no effect on the setting or the relative levels of power. That's just not the sort of things players come up with.
a) Tower made of Iron and can be shrunken down and restored to size with a command word.