Data driven approach, eh? How's this for data: The reams of comments and feedback threads (and their upvotes) telling you this new UI is junk, is hard to use and hard to look at, straight up doesn't work on occasion, and that the old UI was better, and the actual constructive criticism and suggestions by users to alleviate this?
How about the ratio of "like" reactions on each comment stating the above, relative the ratio of your "explanations?"
How about the fact that a large proportion of the users complaining are the same people paying money for site features, and that perhaps their opinions should be considered a bit more carefully than "complaining about change" and being classed as a "vocal minority?" (That's to say nothing of the fact that the vast majority of people probably don't even know the forums even exist, since they're hidden away under the community dropdown, not to mention that they're not exactly intuitively laid out. I'm a smart guy, and it took me a solid few minutes to figure out the right place to post)
Where's this data you speak of? Can we see it? Can you show it to us? Or is it all in the realm of "just trust me, bro?" I could say I have data to prove that Valve will release Half-Life 3 in July of this year, but that doesn't make it true, and unless I actually present it, I should (rightly) be ridiculed.
Someone else further up the thread mentioned that you don't need to change the UI when you change the backend. You don't want to support both the old look and the new look with a toggle, like both reddit and imgur do (not that that's a back-breaking ask)? Fine. Ditch the new look, and point the new code to the old UI assets. Your users don't want this. And since your users are the ones, you know, using the site, you should perhaps listen to the data they're giving you. Especially when a good portion of said users are responsible for footing the bill. We don't care what your "plans" are, we don't care whether it's your pet project or not, we don't care how much was invested into it; if we don't like it, if it makes the site harder to use and uglier to look at, if it renders it literally unusable for those of us with eyesight issues, then we DO. NOT. WANT IT.
This guy gets it. Especially point 5.