On cloudless days it it very accurate because on those days UV strength depends almost entirely on the solar altitude, which can be calculated very accurately from the location, the date and the time.

A sibling comment mentions the ozone layer, but I severely doubt it varies enough to be a source of inaccuracy.