" The numbers that stabilize quickly" is doing an unacceptable amount of work in this tweet. Rather than asserting that, we can readily look at his historic monthly numbers. All of these will be month by month, combining his performance against all pitches - he's had basically 1/2 of a month this year, and that shows up as a point at the end. Showing monthly performance is important because it lets us see the scatter in his historic performance.
This is a good one to start with, because we see that his barrel rate is quite low right now, although not the true lowest of his career. However, we also see that he commonly bounces between 5 and 15%, with a high of 20%, so a single 2 week period at 2.5% is low, but the range is so large that the difference between 2.5% and 5% isn't certainly predictive.
Here is the next one, Exit Velocity. Obviously this is actually down, again it's the second lowest of his career. But, monthly totals varied last year by almost 6 mph, and his best highest exit velocity was associated with his slow start. This is concerning but there's a lot of variation again.
Is Jose Abreu suddenly failing to swing at pitches in the zone? It's slightly low, but it's not unreasonably low compared to his career.
Is he chasing more than normal? His chase rate is higher than his career average, but it's no higher than in any random month throughout his career. Thus, is the phrase "He's swinging at a lot more junk" accurate and well supported, on the grounds that his numbers these first 2 weeks are higher than his career mark? It may be true that it has ticked up relative to his career mark, but it's totally within the realm of monthly performance in recent years.
Here is something that is abnormal. When he is swinging outside the zone, he is making less contact. This is about the only one where I would say his performance is extreme relative to any other month of his career. But, again, the variation helps here - remember this is 2 weeks of performance, so you'd expect that the variance over 2 weeks would be larger than the monthly variance.
Abreu is making less contact than last year on average, but he's making more contact than May of 2021, for example. Here's another case where comparing his 1 month performance to the average over last year is missing the variation he showed last year.
There might be something to him making a little less contact, and perhaps some of the other things such as his low exit velocity stay put and wind up mattering to him. But given the monthly variance, the claim made in that tweet is stretching the data beyond the breaking point, it isn't fair to draw that conclusion from the measurements we have so far.