I don't know if it's a good rating system, but I don't like that it's trademarked, it reminds me of Agon (World Chess Championship 2016).
As long as they don't show the source code, we don't know if it's better than Glicko2, the rating system lichess uses.
In any case it's very unlikely. Glicko2 is very good.
Also, one rating for all time controls? Obviously quite ridiculous.
That USR seems much more like a publicity stunt to use in GCT rather than a serious attempt. From mathematical perspective I don't see a viable model that would give accurate strength over all time controls better than 3 separate ratings.
Considering it's trademarked, it's also likely to be patented.
Amusingly, I'm fairly sure lichess had something similar to this in its early days when blitz, classical and bullet ratings were all pooled into one number. There are significant issues and problems with that method (which I was initially reluctant to change from, but now glad!)
Additionally, the fact it's trademarked isn't really hugely encouraging with regards to St Louis' claims about trying to make chess more open and transparent. I think it's clear, personally, they're trying to fracture FIDE so that we (once again) have more than one dominant chess federation.
The idea of URS seems kind of ridiculous, because firstly there can be no ideal model that is a balance of classical, rapid and blitz and secondly it seems like an attempt to slowly start taking over from FIDE compared to anything else and Im not sure how the idea will work but Im not a huge fan.
If it's as good as they say it is, then we'll all want it. But the announcement you linked is nothing more than promises of wuality and credibility, not any demonstration of them. For now we can only say, "Where's the beef?"
