That first part makes the assumption that every card boosts the damage of it's target in the same way, via their attacks. It doesn't have to be like that, though. For example, if a hypothetical Bole were to give 3 stacks of 'reflects damage equal to 10% of the target's max HP each time they are struck' (and the mit effect for utility/flavor), then it's not better to put on the DPS, but on the tank, since they have more max HP. Ewer could be given an effect that makes it more incentivized to use on a healer over every other class. The option would remain to prioritize using the utility creatively over the 'correct damage choice'. For example, saving a DPS with a well placed Bole would still be possible. And yes, this would cost damage in a way that current AST does not. However, unlike SB AST, which many seem to agree was good and also cool, doing so here would cost only some of the damage. If the stacks were to have no duration for example, lasting until a new card overwrote them or were all spent, then you could recoup over 50% of that 'lost damage' by just waiting for enough raidwides to hit the party to eat those stacks.
I can see why we'd want to segregate the cards into say 3 'only utility' and 3 'only damage'. But then we have the issue we do now: those utility cards potentially won't see much of the light of day outside of Savage. Then we've taken the deck from 6 down to 3
I should clarify, when I say 'Balancefishing', I don't literally mean fishing for the card called 'the Balance', it's more like a phrase I was using to convey the concept of what you described: Get Expand, then fish for the best damage card to spread with that RR effect. Not necessarily Balance, depending on comp, tuning, etc. But in that moment, it was Balance. If the Balance was nerfed, we'd just do the same 'Balancefishing', but with Spear. Or Arrow, whichever was better damage for the group. If the Expand effect was removed, however, then we'd have to completely change up our priorities. Extend was good on some classes, and not so much on others. Empower or whatever it was called (burning a Bole) was better for bursty classes, not so much for sustained classes. We'd have to adapt the RR more based on who we planned to give the card to. But, Extend was so much of a gain for everyone, that it was just 'the best one' in every situation except 'pad the one guy who's going for a 99 log'
If the aim is to remove the 'concept' of Balancefishing, ie redrawing cards you get because you're in search of 'the highest damage card', then every card has to give roughly equal value in damage. It's either that, or some cards are worse but some external factor makes up the difference (eg playing Ewer is bad for your damage, but empowers Div to be stronger, which totals more than 2 Balances would have), which then causes a 'this feels bad' feeling for the player in the moment, when they play the 'bad card'. I just think that the main factor in why AOE Balance was so busted, was the AOE part, not the card, so that's what I'd use to solve the problem: keeping the cards single target.
It does a little, yes. But we're looking at it from a perspective of 'today', rather than in contrast to 'what people want from the past'. As above, compared to the past this would be a godsend. Losing only half a (singletarget) Balance damage from playing a player-saving Bole instead of a whole one? I'd assume people would be a lot more open to that if they were asked back then. The issue with Div being tied to Seals (as it was in 5.X) was that it felt exceptionally bad if you did not get 3 Seals, because it was a punishment that is multiplied by the size of the party. I don't see how it's too bad a problem for people to have the 'meaningful choice' between using the utility more creatively, or just full-send the damage as intended. If we say 'oh we can't have MP on the Ewer because people won't use it for MP, it'll just go on whoever gives the most damage (under my tyranny, that'd be a healer)' then we end up where we are now: every card is exactly the same, because we didn't allow for any potential for 'it does less damage'
Actually, I take that back: using a Bole on a DPS and only getting half effectiveness (as described above) is possible now: Play a Bole on a melee. It's 3% instead of 6%, halved effectiveness. Sometimes you get screwed by RNG, and draw a ton of ranged cards in a melee heavy team, or vice versa, and have to decide to either play a scuffed half-card, or hold it until after the first cards fall off and it's no longer bursting time. So if people were able to make the choice to play a card for it's utility in this hypothetical, and still reap half (or more) of it's effect despite 'playing it wrong' (eg a Ewer on a res'd caster to help their MP til next Lucid), I'm not sure I see how that's a bad thing. Especially when they are choosing to do so for the additional effect like Mit or MP or movespeed or whatever, whereas at the minute we do it because 'well there's no other player to put it on, and the 2min meta means that most of every class's damage is now in that bursty 15-20s window, so I have to even though it feels awful'