The page has no discussion of prior art like PCG[1], and no results from standard tests like Diehard[2] or TestU1[3].

It might be good, but I'd stick with proven alternatives.

[1]: https://www.pcg-random.org/

[2]: https://en.wikipedia.org/wiki/Diehard_tests

[3]: https://simul.iro.umontreal.ca/testu01/tu01.html