Tracking metrics on an individual user and hard-coding them into the source is the cringiest, most insecure, egotistical thing I've ever seen

This looks like an experiment to measure the impact on code changes on the distribution of tweets of different user groups. (comment down below confirms it). Elon is a good candidate for the test because he is the most followed user, controversial and can provide consent to collection.

My guy, according to twitters own admission in the same comment you are talking about, it clearly says in t he last sentence, "if any changes negatively affects these groups one over the other , those changes will not be pushed" (only elon is included as a separate category, the rest of the categories are groups of individuals, which makes no sense as he should be already part of VITS), so yes, since there are media reports that reported that he specifically fired a developer for showing how his popularity is in general decline after the acquisition in google trends, there is a very high chance that this is done under his command during the same day he fired that engineer.

So this is either a big april fool prank that is executed so poorly and hence missing the point of Aprils fools prank.
Or
Don't stan someone to this extreme degree that you are ignoring the evidence at hand.