This is an interesting paradox... So, from what I'm reading here, people seems to be of the generalized opinion to "wait". So that means DDR5 modules will continue to be injected into the market, but there's going to be a point where retailers just won't buy them anymore (over-stock and stuff) to sell current ones. General public will definitely won't know the difference between DDR4 and 5, so they'll only notice how much more expensive DDR5 is and see if they can get a DDR4 platform, or it'll come down to OEM's to price stuff competitively. Then again, major OEMs don't care about AMD anyway.
I'm thinking that AMD will have to time this one very well and/or push Samsung or Micron to get better modules out there for AMD to justify the gains. Which also brings me to think about that: "gains". At what point do we (yes, me included) stop whining about the price and look at the actual performance benefit? I personally do not think DDR4 is currently holding back Alder Lake horribly, but if you look at the hard data for similarly timed RAM (DDR4-3200 CL16 vs DDR5-6400 CL30) it is undeniable that Alder Lake gets a massive boost in performance in most things, games included. Then you have AMD's 3D-V cache which should massively help to hide the horrible latencies from DDR5's lower end kits (or so I predict?). Maybe worth investigating with the 5800X3D? Although, when the IF is clocked higher, it does give the whole CPU a massive boost, so that brings me to another angle: as long as you can decouple the IF (FCLK) from the RAM clock, then AM5 may be able to get away with very good performance using crappy kits?
Hm... A lot to think about and speculate, but interesting to check now that there's a "sample" of what AM5 could become thanks to the 5800X3D.