Long time Amazon seller/brand here (and ex-academic/game theorist) -- it's a little more complicated if you think about the various parties' incentives, but yeah we have a half-assed market equilibrium that sucks for everyone.

Amazon as the marketplace facilitator has an interesting job to run:

1) It wants customers to get exactly what they want (as measured by the minimal amount of clicking/scrolling/time it takes for a customer to hit the landing page -> complete shopping). This is mostly customer-centric.

2) It wants to promote as much competition as it can between products and brands so that no seller dominates the market. This is... mildly customer-centric so big sellers can't rest on their laurels, and mostly self-serving (so no single seller/brand can gain enough power to threaten Amazon).

3) Amazon has to strike a balance of #1 and #2 so that there's some opportunity for new entrants to enter any product space. Back in the days they structured search so that some percentage of results would be new sellers and products, and new products would get some amount of sessions to prove they convert within a certain statistical distribution to land a certain rank (BSR). Note the rise of alphabet soup Amazon brands is partially because they allowed for disposable brands that could be black-hatted, and partially because for a few years these sellers were just generating tons of new brands to share new SKU real estate on search results (so when you searched for "small kitchen colander" or what not, you'd find the same product being sold by 10 sockpuppet sellers occupying 10 search spaces, taking result space away from the other sellers).

But what if they can get everyone to partially pay for #3? Enter ads, where basically new and old brands pay to play for eyeballs on Amazon.

So if you think about it, Amazon is a search problem where you wanna push results such that the expected value of the results pushed yield the highest expected value, something along the lines of: Expected Value (EV) of your product = [your sales price] x [your product's conversion rate in the past X days] x [some confidence level of said conversion rate] - [returns] + [whatever other Amz secret sauce] + [Amazon PPC]

Amazon can get people to pay for their eyeballs, swap some seller blackhat tactics for others, and more or less keep OK results. So that's Amazon's end.

As an established seller -- I know my product converts at X%, so as long as the math works out where I spend $Y/click, expect it to convert at X%, and my profit Z > Y * X it's what I do (and the tax I pay), and I move that up and down depending on my inventory situation. If possible, yes, I would pass some of that down to price raises as long as the market's willing to bear it, so I can make a minimum margin requirement off my Amazon channel (which is the lowest, compared to my D2C website or through offline B&M channels, but I do so because my biggest volume is on-Amazon).

As a customer... yeah the default search results are "good enough" for the 80-20 crowd; the convenience, ease-to-checkout, and 1-2 day shipping wins them over. As a "nerdy" customer who likes researching products to get the best one, I used to research via online reviews and reddit (most of which are gamified in their own ways)... but now I use AI (ChatGPT, *not* Amazon's Alexa Shopping because it's not great) and mostly call it a day.