If I had to guess, I’d say the turnover stats only count full employees, and are therefore a reflection of their “trial period” hiring policy more than anything else.

If they were deliberately manipulating the stats, that would come out quickly with the level of scrutiny they're facing right now. Also, wholeheartedly agree with a trial period. They're good for the company obviously for lots of reasons, but a less obvious one is rooting out cultural fits and problem people like this. Also, I think this is better for the employee as well - it's much better for mental health to have a clean break than it is to spend the next 6 months going through "performance improvement plans" and such.