Interesting take here: https://www.techpowerup.com/review/intel-10900k-vs-amd-5900x-gaming-performance/
Though no mention of dual rank memory or not. Perhaps Ampere does see some benefit from having PCIe 4.0 vs 3.0 on the Intel side too since it's the only way they were able to get AMD to "win" the FPS battle.
It does make things more complex for the average system builder, since it would appear certain combinations of GPU/RAM perform better than others, it's not as simple as Intel where you can slap it together and (generally) get top tier performance.
Thanks for the link. That was a great article. Coupled with the memory articles from Gamers Nexus and Hardware Unboxed it is becoming apparent that it will be difficult to cross compare results in reviews. I know that all the reviewers and reviews I have been reading are pretty detailed on their test platforms and methodologies, we can see that variation in platforms can create some very different results!
I guess a main take away is that each review becomes its own encapsulated data point. The reviewer's comparison and deltas have to be viewed more from a percentage point to point delta vice raw FPS numbers or results. It certainly makes figuring things out that much harder.