So, casting tablets... tricky issue. While some people feel that the possibility to case some spells without having to actually re-invent them makes the game more accessible, as it stands right now casting tablets are extremely "powerful" - as in that they raise the question why magi would even bother with learning some specific spells as well as why there is no abundance of high-level general spells that see a lot of use (Aegis of the Hearth, circular wards, etc.), given that casting tablets can be copied like lab texts. Also casting tablets don't require any extra time for casting and thus are really infringing on the playing field of formulaic magic.
I had an interesting idea yesterday: Casting a spell you don't know is spontaneous magic, so why not handle it that way and have casting tablets give a bonus to that? Using the rules for ceremonial magic seems like the best idea, given that you have to cross-reference the instructions on the tablet in order to benefit from the bonus. Therefore casting tablets take longer to use and don't make sense for any spell anymore. Casting a Pilum of Fire that requires an hour of preparation? Unlikely. Casting Purification of the Festering Wound? Casting tablets sure come in handy there.
The only question to me is what the bonus should be. Several ideas:
+1 per magnitude of the spell. Works nice and simple mechanically, but raises the question of the in-game logic of it. (It's a plain 10% reduction of the base ease factor, so it might be justified as plainly being "easier" to cast.)
A flat +5 bonus.
A bonus that is set while writing the lab text, making it harder to write the higher the bonus is (+1 per 1/2/3/5 levels of the text over the level of the spell?). Probably open to abuse.
The sum of the Artes Liberales plus Philosophiae of the writing magus. (Might also serve as a cap for the previous option.)
Another option that might be added is that casting tablets reduce the time for ceremonial casting from 15 minutes per magnitude to 5 minutes per magnitude.
Casting from tablets is pretty botch-prone - which makes casting rituals from tablets verrry dangerous, as well as casting from tablets in non-magical (or even unfamiliar magical) auras.
Then, penetration is designed into the tablet - so spells whose effect depends on penetration are often useless when cast from tablets.
IME casting tablets, as they are now, work as stop-gaps for young magi, or for groups with only a few magi. And everybody - most of all the designated user of particularly useful casting tablets (like Piercing the Faerie Veil ... ) - is happy to drop them as soon as possible.
10% chance to check for the botch...
But if you have to check...then you are very likely to botch. Casting that Aegis from a tablet? Say a 7th magnitude Aegis, that's 8 botch dice to check... 1 for the spell and 1 for each pawn of vis. Wait, your CT wouldn't be high enough, so you need to commune? Add another botch die par participant in the Communion. Finally, someone in the Communion botches, you botch. Let's say that there are 2 participants in the communion. 10 botch dice. Pretty sure you're going to botch...
Once you do botch, add 15 warping points, 1 for the spell, 7 for the vis, and 7 for the magnitude of the Aegis. Congratulations, your brand new magus has achieved a Warping score of 2! Enjoy the Twilight.
So, while the chance of having to check for a botch is 10%, once you get past that threshold, and have to check for botch, you're pretty much assured, for all of the high level, useful and interesting ritual spells, that you're going to botch. And it's not pretty.
As you mention yourself later on, allowing casting tablets to spontaneously cast ritual magic might be an option (maybe at 15 minutes per magnitude?).
Since a fatiguing spontaneous spell of the first magnitude has a base ease factor of 10, a +1 does actually account for 10% of that, if I'm not mistaken.
I like the option as a storyguide, since it gives me the means to introduce spells that no one else might be bothered to actually learn in order to cast them.
According to my math that's a net 5.7% chance of a botch for every casting of the spell. Given that it's only cast once a year, some might consider this an acceptable risk. (Also a single botch die coming up as 0 doesn't necessarily mean catastrophic consequences.)
Now you're just trying to create an unfavorable scenario. Apart from the fact that there is hardly any need for a communion when using casting tablets, since the spell only fails if you miss the ease factor by 31 (!) or more (which in the case of your Aegis example would mean a casting total of 4 or less after rolling), in all my years of Ars Magica gaming I have yet to witness a single Wizard's Communion being cast. None of the players have ever bothered learning it, restricting themselves to spells they can actually cast on their own.
Yes, but that is not RAW, so not relevant at that point
That's going to be a fight over semantics. And I have a fever, so I can't be arsed at the moment.
I've generally found that if the need is there, someone will bother to learn any given spell.
On the other hand, I currently have a character in a (spring-ish) covenant with no Aegis - lot's of stories there!
I'm not going into probablities, but even a single 0 on a botch check for a ritual is catastrophic. It means you botch, which means you get warping from the botched spell, and for all the vis used in the ritual. It's bad. Really, it is.
I did use a 35th level Aegis, because of Conjuring the Mystic Tower, which would be a spell I would expect to be on a Casting Tablet, as it isn't very useful to learn/cast beyond the first time. Young magi might be willing to accept the risk of Warping for a few years, but eventually it will get old, and eventually the odds will catch up to them, and they will botch, and boom. There is a very real Risk/Reward profile here. I see very few magi casting spells regularly from tablets because of the dangers.
You need to review the rules of botching. When one botches, the total is 0, without any additions. So if someone botches a 35th level spell, they have missed the ease factor by more than 31.
As to why your players don't use Wizard's Communion, it really only becomes necessary to push casting totals to achieve a higher penetration on wards and Aegiedes of the Hearth. If you don't play with wards that penetrate, per RAW, then you have less of a need of Communion. Taking it a step further, it makes casting/learning the Aegis ritual within reach of almost anyone: Int 3, Rego 5, Vim 5, MT 4, Aura 5 for a total of 22, boom 20th level Aegis from the lab text learned.
A botch always means that your total for the roll becomes zero - which for the 7th magnitude Aegis cast with a tablet means a Casting Total - Spell Level of -31 or less: 7 Warping Points on top of the Warping Points for botch dice, and check for Twilight. That Twilight is pretty assured, and nets you another simple die of Warping Points.
I'm well aware of how botching works. I'm not saying there is no risk at all, just that it's rather small compared to the potential benefits. Also this risk mostly applies to ritual spells - high level non-ritual spells (like "kill someone you have an Arcane Connection to" (PeCo 50), or maybe even a high level Wizard's Communion) would turn out to be highly useful at a minimum risk.
As for my sagas, there never was any need for the Aegis to be cast from tablets, since there was always one player more than willing to actually learn it (the saga I'm running will soon have an Aegis of eight magnitude) - it's a spell you regularly use after all.
That high level PeCo kill somebody spell needs to have penetration built into it, so keep that in mind, and when you're doing that, you need to think about who that person was, and why they made a casting tablet. How they had a monster PeCo Lab total, and how they mastered the spell... Maybe it's a bad example, though.
Casting Tablets are typically for some utility purpose, or one off purpose. I play my magi as generally risk averse, except in the areas where they seek out danger. Combat magi do combat, but other magical activities they try and avoid botches. A magus who researches only risks the warping from research, they too avoid opportunities to botch. And as they age, this tendency only increases. If they use a spontaneous effect frequently, make a spell, if they expect to use a spell in a foreign aura, master it to minimize botch dice. I'd only use a casting tablet for a one time, or very infrequent effect, because the risks of botching are just too great. One Shot's Piercing the Faerie Veil is a perfect example. InVi seems to be a rather low priority combination for a number of magi fresh out of gauntlet, so that casting tablet has some utility, but as soon as the PC can learn that spell, I'd be learning the spell, especially as the necessity to use it increases.
Casting Tablets, IMO are a crutch, and nothing more. If you rely upon them after you're leg has healed...probably going to turn out badly for you...
Look, you're starting with what I take as a false premise: that Casting Tablets are too easy to use. I disagree with it, and don't see any need to make them harder to use. If you want them to be harder, make them harder...