We discussed this at length a couple of years ago. There are too many implicitly static container spells in published books for this to be viable. (And too many implicitly dynamic spells to make everything static.)
Yes, illusions feel as though they should be dynamic. That was one of the considerations pushing me towards generalising the distinction. The explicit designation of Group as static (which I suspect is a large part of the reason why so many people intuitively feel it should be, even if they no longer remember reading it) is something I don't want to overrule, and then considerations of not adding yet more weird exceptions push me towards static Individuals.
Still, making Groups potentially dynamic (by removing that passage) is a "widening" of the possibilities of Hermetic magic that requires no additional errata, and makes possible some existing spells that would otherwise fail to work as expected. For example, many large-scale weather spells are T:Group (each member being one cloud, one gust of wind etc.). Similarly, many large-scale Aquam spells are T:Group (e.g. "a large river, where there are several different currents"). For those lasting more than D:Diam, individual elements come and go, so you need a dynamic group if you want the spell to avoid developing "holes" that eventually devour its entire fabric. I would add that a spell affecting an army (e.g. making it invisible, nigh-inexhaustible, or berserk) feels more mythically appropriate as dynamic, rather than static: new recruits who join should be included, old hands who leave should be excluded.
In general, I would strongly favour every Target being potentially either static or dynamic, for at least two reasons. First, because it makes the rules simpler, removing the necessity of distinguishing between "object" and "container"; a distinction that is not always easy to make for new Targets introduced in later supplements. Second, because for many illusions, as you mentioned, it makes much more sense for e.g. an Individual Target to be dynamic.
Finally, I think that it makes sense for Part to be dynamic too, in some contexts. If I change the colour of my shield grog's hair with a T:Part, D:Moon spell (so that she can still punch those who offend me without having to overcome their magic resistance) I definitely want a dynamic T:Part so that regrowth is not visible, or she'll punch me instead!
I think that Individuals should be, indeed, static, in regards to illusions at least. I think we want to be able to throw floor/ink at an invisible person and have it reveal them as the default. Or the valuable diamond "floating in the air" after grabbed by an invisible person (unless the person takes efforts to hide it, putting it inside a pocket for example). That might be just my personal opinion.
This might not have a single satisfactory answer that works across all kinds of spells. I don't like the idea of casting a R:Personal ward against flames on myself, losing my hat and having it continue to be impervious to fire (static). I don't like the idea of grabbing a random hat and having it suddenly be warded either (dynamic). I'd rather have something like
the spell ends for things that cease to be part of an Individual (an invisible hat becomes visible if dropped), but if things become part of an Individual the spell does not extends to them (the hat does not become invisible again if picked after it is dropped, nor any other item grabbed after the spell is cast).
This also means that if I use a MuCo(An) spell to change myself and my clothes into a wolf and then put on a leather hat it won't disapear, becoming part of my fur.
Then again, maybe this isn't something I'd want for every T:Individual spell. There might be corner cases. But I can't think of any at the moment.
Do they? An air or water current is an abstract concept anyway. It does not necessarily have a beggining or an end. There is no implicit concept of "comming and going", or at least, no need for one.
As for clouds, well, we get a bit Shippy of Theseus here, but as I see the clouds can change their shape and (within a certain limit) size without changing their Individuality. In a magical storm the clouds change, but they don't really change. This would be a dead giveaway of a supernatural storm for anyone that actually knows the trade ("I'm telling you, those clouds ain't natural!"), and has an appropriate mythic feel IMO.
(It's also in line with other spells. Any craftsman worth their salt can tell that an object created by magic is somewhat unnatural, and any shepperd will be triggered if all the ritually created sheep seem like exact copies of each other, or if them start to behave as if mind controlled under a ReAn spell.)
I'm not sure that widening the possibilities of Hermetic magic in something that is supposed to be errata is really a good idea. There is a definite risk of power creep.
In the abstract, I would agree. But in this concrete case, do you really think that allowing dynamic groups would create any significant "power creep", particularly compared with how the core rules have been "expanded" by subsequent supplements?
Furthermore, this slight "widening" (in addition to making things simpler and clearer) in fact allows the RAW to make "legitimate" some spells that, on careful inspection, are not quite legitimate according to the current RAW. Rafael Bessoni tries to fuzzy this a bit by saying:
But the fact is that, in many bodies of water, a current is not "abstract": it is a very real thing, real enough to kill you. If you live by a large river running across fairly flat lands you know that when heavy rains arrive, it creates new sidestreams. If you live by the sea, you know very well of tidal currents coming and going twice a day. This are all Individuals by the RAW, that join and leave the main Group. I argue that spells acting on such Group targets for D:Sun or longer are already effectively using dynamic groups.
I think in the context of the Aquam guidelines (and in my post) a "current" is meant as "moving water" rather than "the movement of water". So it does encompass actual water (even if the water changes over time). And it's indeed an example of an Individual of water, not a Part; hence a body of water sufficiently large/complex to have different currents is always a Group and never an Individual, explicitly by the RAW.
Fair enough; but then spells that already implicitly use dynamic groups (if they work as one expects them to work) should see some errata ... or at least clarification saying that they do nor work as one naturally expects them to work.
That said, just for curiosity's sake: do you see any example of "excessive power" in allowing dynamic groups? To me they just appear to enable appropriately mythic effects, including "classic" effects that were part of Ars Magica since the 2nd edition. If anything, it's the static groups that put game balance at risk (that's why at my gaming table we are strongly considering the option of saying: every Target is dynamic, full stop).
Let me add one last thing. I really commend this clarification/clean up effort. It's very useful, and long due! However, it seems to me it's uncovering some very significant, fundamental complexity in the structure of the game that had been so far "shoved under the rug". In the light of this, I think the value of simple solutions is growing with each additional clarification/erratum made.
Are there any such spells? I mean, the current rules are pretty explicit that Groups are static, so I would expect the vast majority of spells to follow that.
I might have explained myself poorly. You can say "there's a strong current at this point of the river". In that sense, it's not abstract indeed. But where does the current begin? Where does it end? This is fuzzy. A current is not simply "a definite and discrete amount of water moving". It's a phenomenon rather than a discrete object.
Think of a stream (a single current). Water may enter and leave the stream, but the stream itself doesn't change. The same is true of a river (a group of currents). This isn't the same as targeting a discrete amount of water (a puddle, a lake, the water inside a cup).
This all to say that (air and water) currents work perfectly fine as static targets, because the targeted individual itself is preserved. Just as a horse may eat, drink and poop, and the simple fact that exchange of matter occurs does not make it a different horse at each passing second.
I can definitely see that. Unfortunately, I am getting the distinct impression that the changes required are too big for anything short of a new edition, but not exciting enough to justify a new edition. So, clarifications that can be made without creating a new edition will be, and some of them will be fudges, and not as clean as we would like.