Ryzens are dynamic cpus. Not static like Intels. So they'll boost according to need and ability. So in some instances you'll get max turbo on a few cores, or maybe partial turbo on a few cores and a higher or lower rate on others.
This kinda puts a 100MHz difference in the pointless catagory. Baseclock speeds are basically only relevant to testing process where exact speeds are required as a baseline. Otherwise there's so much variance at any given point you'd not see a difference at all.
Which makes the IPC differences between generations of greater value, core workloads etc. If you take CSGO, that uses only 2 threads. With a 2700x - 5600x there's an almost 40% difference in IPC. Put at the same speeds, that's significant amount of instructions processed. If the 5600x was at 3.0GHz, the 2700x would need to be closer to 4.2GHz to make up that difference in instruction count.
100MHz is chump-change in comparison.
With current gaming loads, even involving discord or YouTube or streaming, you are still looking at an average of @ 8 threads used. Granted a core is always stronger than a thread, so hyperthreading does lose out a little there, but that's really the only reason some games get a distinct advantage on the high core games, while others show very little difference between a 5900x and a 5600x.
Personally, I think the best cpu is the 5800x. Reviews don't like it because of its price point, but it does everything and anything Well. Not the best for production, not the best for gaming, but close enough in both to the best overall that it's price is justified. For pure gaming, the 5600x is the Ryzen to beat. Right on the heels of its bigger brothers, at a much reduced initial cost and cooling cost.
I like the R9's, but if you don't have use for its core count, or anywhere close, it's just too expensive to justify the small fps gains that you won't see anyway.