So I think the best strat here is for the users who did get the new layout to just stop using the desktop version of the site for a while, like a week or a month or however long their 'experiment' is supposed to last, while the users who didn't get the new layout should keep using the desktop version like normal or, perhaps, use it even more than usual.
My guess is that they're doing basic A/B testing on the new layout to see if it would boost engagement: the userbase is split roughly 50/50 between the 2 versions and they are going to be comparing the engagement data between the 2 groups of users to see if it's worth it switching everyone to the new layout or not.
Basically, if you got the new layout and don't like it - don't use it. If engagement metrics of group B (new layout) are lower than those of group A (no change), the experiment will be considered a failure and they will have to reverse the change.
If your tumblr suddenly looks like twitter - it's a sign to log off and go touch some grass! (or just use the mobile app since that engagement data isn't relevant to this particular experiment)
Don't just not use it, send feedback too!
There's a "contact us" option to send feedback about features being launched. GIVE FEEDBACK IN THE APPROPRIATE SPACE!
Not to "As a professional UX researcher" on this thread, but yeah, as a professional UX researcher, now is the exact time to provide clear (but kind!! the poor UX team is usually not responsible for these decisions) feedback on what your thoughts are in regards to this change.
In my job, if I were doing an AB test on a site layout and every person I interviewed said "I hate it, it looks like knockoff twitter, please put the old one back" then I would be very excited to include a nice little bullet point in my report that says "[x] number of participants disliked the new layout :)"
