Aethergems are significantly larger than magems. Especially the smaller ones. As far as I can tell, they actually use the same generator component on all of the aethergems, just with different magems on them. Point being, there is good evidence to suggest that if you just saw an AAA agem in half you will get a generator and a battery of roughly the same size. Sticking a whole lot of generators onto a battery will get something huge. Unless magems are actually tiny?
Calling Aethergems capacitors is misguided. We have capacitors. They are called mage gems. They do just fine in that role. Capacitors do not produce energy, they just store it. We also have generators, the aethergems, which are doing rather poorly because they are trying, badly, to replace our capacitors while being our only generators. As is plainly obvious, the only aethergems that are worth anything are the smallest one, because we can use magems and AAAthgems better for any task that would use AAthgems. When using Aethergems, the only thing that we care about is generation ability, because storage capacity can always be made up with magems. In modern devices, you have generators and capacitors, and the two are separate. You can have a big generator and a little capacitor bank or a little capacitor bank and a large generator bank. It depends upon whether you want a frequent use(powerful generator) or occasional use(weak generator) and whether you want strenuous applications(large capacity) or weak applications(small capacity). Unless we just stumble blindly into the perfect ratio for a given application, an aethergem is going to have a very poor capacitor to generator ratio. That is just the only outcome of having your ratios fixed like that. We currently have capacitors and generator/capacitor hybrids that really don't come out working as either particularly well. I get that you really want your all-in-one does-everything universal-power-unit but that is, quite simply, a bad way of doing things. Adding some magems and switching the bulk of it over to gemerators would be as simple as any of the other proposals and produce a purely superior outcome. You are basically trying to make my gemerators except instead of explaining how this can actually make a better system and trying to improve the generation facilities in the process, you are just saying that, in spite of all the evidence to the contrary, if you just shrink the capacity down to the point that it ceases to be credible, the aethergems suddenly become capable of being absolutely tiny with exactly the same power output, as though, perhaps, as I have been saying all along(although we disagree as to whether the zero capacity aethergems would really be that small), binding them to a capacitor is holding them back from their full potential, but then you go and ruin it by, surprisingly enough, binding it to a capacitor.
Separating generators from capacitors is the only sensible way forward if we pursue this technology. The silly way forward is to have some sort of nebulous adjustable capacitor to generator ratio device that does whatever is asked of it because magic. The stupid way forward is to perpetually insist that we must have fixed ratios of capacitors stuck onto all of our generators. We just get a separate capcitor bank which we have had for ages, plug it into the new generator bank that can be attuned to precisely the generation potential that we require for the given role, and not try to turn them into the same thing is a tiny range of fixed models that don;t really work exactly right for anything ever.
So, in short, we replace the storage facility with magems, which I already specified in my proposal. And because my system is objectively superior, we only need two capacitor banks instead of four, unless there is a limit to how quickly they can charge that I have missed/forgotten? So just be separating them sensibly we cut out half of our magems, presumably reducing the volume by about a quarter.
I also suggest that we will not, in fact, get AAAthgem->atharray upgrades automatically. I cite that we did not get magem->aethergem upgrades on account of the dimensions changing, and suggest that the array is, at the very least, going to have a different shape to it, even if it impossibly retains the same size.
Gemerators: Separate the generator from the capacitor and increase the generation ability, then take the obvious and seemingly simple steps to update those applications that seem to be easy to update.
Aethergem Array: separate the generator from the capacitor then put a whole lot of generators onto one capacitor. Then use these to replace the old model because they are exactly the same size with vastly improved performance even though all we did was change what they were plugged into.
Important notes that were omitted:
: That the "generating" component of the aethergem is actually getting smaller as the capacitor/battery component gets smaller even though they all generate at exactly the same rate(citation?) and that we can continue to benefit from what is, obviously, a purely beneficial effect of reducing until it is no longer a capacitor at all, down to effectively zero capacity even though we have, thus far, only managed to go as for as the AAA version which is presumably, a long long away away from practically nothing at all.
: That all of the effort will be applied to maintaining compatibility with the existing array. If sacrifices have to be made, then efficiency can bite the bullet and all sacrifices must be made for 1st: must fit into the old physical slot because we don't want to burn another revision on this; 2nd: must maintain enough capacity for the aeroplane's engines to fly; 3rd, must have greater generation because without that there is no point.
The array is objectively inferior. It makes no stated effort to improve performance, it just assumes that the reader will get the point and understand that the process of a generator plugged into a smaller battery generates at the same rate will continue endlessly. AND that the generator components are actually different, at all, across the different sizes, and it is not just the capacitor components which are larger or smaller. It is entirely possible that the effect which is relied upon doesn't actually exist. But that doesn't matter because it is not specified, it just has an inspired piece of stage magician work of sawing an aethergem in half and gluing a dozen of one half back onto one of the other half and the result is exactly the same size as the original.
Ugh, the fundamental problem is that the aethergem makes no effort to explain how it works, or even that it does work, and assumes that it will be useful even though the prior aethergem which was largely the same level of disaster was not useful because surprisingly enough when you add and remove stuff you get a different size and shape.
The gemerator and the array work in exactly the same way. They both reduce the storage capacity to nothing. The difference is that the gemerator makes some attempt to improve performance and explains how to adapt to the new technology while the array insists on bulking everything back together in the latest episode of "we are not allowed to have independent generators for some reason".
If the array stopped at just "make an array of these the same size as an aethergem but 100% generator. Then the two designs would be very nearly identical, aside from the array not actually explaining the crucial part of why it could possibly work and making it apparent that the foundation of the theory is a bit wonky at best.
It is the vultures all over again. I put up a decent design, then it is overshadowed by an objectively inferior knockoff. I admit that most of the time designs have their own flavour, "objectively superior" is not generally applicable, but falcons were more ambition and less effective than vultures, and the aethergem array is less founded and provides inferior performance when compared to the gemerator, but both birds were birds and both generators are generators, just one is better than the other, and was proposed first, and lost the vote, because the real magic isn't in the wands... I have no objection to being outdone in the same field by something that looks to have been based upon my idea, it is just irritating when it is straight-up worse and still gets more votes.
But yes, what we really need is pretty much anything other than an aeroplane that can't hope to fight against their air power. It was once said that we shouldn't do wind magic when they are already better at it than us. Clearly that was false, but doing air-forces when they completely control not just the airspace, but the whole world up there, is not terribly wise.