PaulAlcorn :
I have marketing materials (reviewers guides, press releases, slides from briefings, etc.) that say, specifically and repetitively, that XFR is only on X SKUs.
Technically, I must agree with JocPro. In practice though, XFR has a full 100MHz (some say, up to 200 MHz in extremely rare cases) range on 'X' parts, while it's limited to 50 MHz on non-X - thus, nothing to write home about.
It would have been interesting to take the boxed cooler into account, too: from what I could see, it's good up to 3.8 GHz, allowing to overclock the chip on the cheap and thus completely destroying the 7600K on the performance-price ratio as it goes from a 20 to 60 bucks difference.
Last note, you keep repeating that Intel has an IPC advantage over Ryzen; personally, it's more of a software optimization edge. Taking a single-threaded, heavily-optimized but using only "basic" instruction set software like Lame, and comparing the 4 GHz Ryzen with 5 GHz Kaby Lake, the latter has 20% better clock speed but doesn't get 20% more performance over Ryzen. For tasks like MP3 compression that easily take place essentially in-cache, the rest of the system can be quite easily ignored.
From what I could find, Ryzen is good to excellent in basic x86, FPU, SSE1-3, but has some progress to do on SSE4.x. AES is probably the most glaring performance dark spot on the chip, being roughly half as fast as Kaby...
But then, I've yet to see anyone finding out the actual strengths and weaknesses of the processor; Tom's used to do that, but it's been a while.