Except if you read the blurb they point out that it looks like an engineering sample and the reviewer states they didn't get it from AMD. Meaning as always happened, someone who had engineering samples sold theirs on some forum or to a friend, reviewers aren't supposed to do it but they do, I can attest having bought a... damn, I honestly can't remember. I think I got an engineering FX chip of some kind back in the ath 64 days, off a reviewer from a UK site.
So they highlighted reasons it wouldn't be completely representative and posted the news, lots of other places are posting leaks. To me, this suggests even ES with bad ram is basically matching 7700k in 1080p gaming... I'm not sure how that is bad news. Sure people who don't read it's an ES will use it as troll bait, but they'll use anything as troll bait anyway.
This news basically says, worst case scenario, it matches 7700k in gaming, beats Broadwell-e and kills the 7700k in everything else.
Honestly Cinebench single thread... I get it's a nice measure of single core IPC but in reality who does such work with one core now, it's moved beyond the point of mattering.
When IPC is good enough more cores wins, massively and Zen EASILY has enough IPC to push it firmly into the more cores wins area.