I'm not convinced that Amazon needs to change anything. There's little threat to their virtual monopoly in ebooks. If author A stops advertising and their sales drop, surely it just means that the buyers will have bought books buy author B or C. I've not seen any evidence that their advertising rates have fallen, so there must still be sufficient buyers of the advertising slots.
Amazon's not in the online written content market and is doing what it can in visual content - films, TV - which it presumably sees as higher growth. I think that's more the focus higher up the food chain than KDP management.
What Amazon offers may not be great for authors, especially indies, but is it really benefitting from the huge number of books being self-published every year? Wouldn't it be just as profitable if the number halved? Maybe even slightly more profitable if the costs of adding new books were halved.
As is all too often the case, we need data we don't have to really address these questions, so some of my response is going to be speculative.
If Seattle got hit by a nuclear attack, would Amazon survive? Certainly. But depending on how good Amazon's insurance is, it would have to spend time and resources rebuilding its headquarters. Amazon would probably be better off if Seattle didn't get nuked.
While nothing in indie publishing is going to be as dramatic as a nuclear attack, between sales, KU revenues, and AMS revenues, indies collectively bring in billions of dollars, though we can't know the exact number. Amazon is hardly going to go broke even if indie publishing ceases to exist tomorrow, but it will lose money. Amazon doesn't "need" to do anything, but it could better maximize its own profits if it took a few common-sense suggestions.
It isn't necessarily true that everything in the Amazon ecosystem is a zero-sum game. When I shop (and I think this is true for a lot of people), I look for things that appeal to me. I don't go in thinking to myself that I'm going to buy four books today. I go in looking for something I'm going to really want to read. Maybe author A in your example would have been a perfect fit. But because author A stopped advertising, I didn't see a book I would have bought for sure (which might have led to more purchases in the same series or from the same author). Maybe I'll find something else I like. Maybe I won't. But I'm not automatically going to throw the same number of books into my shopping cart regardless of what I see. That's one thing I can be sure of.
And no, the advertising rates haven't fallen, but all advertisers are not created equal (or else Amazon's ad placement system is totally broken). I keep seeing things in which I have no interest and/or which I can't figure out why Amazon is showing to me. I also see the same ads over and over, even though I keep passing on them. (Well, that probably is Amazon, but the weird selections could be poor keywording or other author errors.) And since Amazon seems to be making visibility harder and harder to come by, as well as not doing enough to keep the ads as effective as they used to be, it's increasingly likely that people will drop out.
Who drops out will make a big difference. Some of Amazon's moves, like eliminating also-boughts, detrimentally affect everyone, whether they contribute much to the ecosystem or not. Yeah, Amazon could probably do with a lot fewer indie books and still be fine. But there's no guarantee that all these shenanigans won't shut down some of Zon's biggest selling authors and leave them with hordes of inexperienced newbies instead. Some newbies will be brilliant, but that would probably mean smart enough to see a losing proposition when it's right in front of them. It's possible Amazon could end up with a much inferior indie author pool over time.
This is mostly basic economics. Why do we have such a bad teacher shortage in the US? Teachers are underpaid for what they are asked to do (a job that keeps growing). Fewer people train for it. And of those who do train, a large number leave the profession within five years. Some of the people left are desperation hires who aren't as strong as the people they're replacing.
Of course, writing is very different from teaching, but similar economics apply. Lower rewards and prospects equal fewer people entering the field. The tendency will be for the most talented people to seek other ways to express their talent. There will still be some people who will continue because, as Stephen King suggested, they will write or die. But how many of those are there? And even their writing time may be limited as it becomes harder and harder to ditch the day job. As the population of indie authors tends toward the low end of the spectrum, Amazon will lose more sales. And fewer and fewer people will be able to afford AMS ads. There's a limit to how many people can sop up that real estate and to what degree.
Does Amazon need to do anything? No. Should Amazon do something? Yes.