I read somewhere (forgot where) that proto uralic may be constructed with having a base 6 number system. Is this bullshit or something to take seriously?
Basically bullshit. There's zero evidence whatsoever for 6 as a base, i.e. for expressions to the effect of 6+1, 6+1, 6×2, 6×3, 6×6… The actually published hypotheses that I'm aware of only makes any argument for a limited number system with unique terms ending at 6, supposedly with no firm evidence for how anything larger would have been expressed. The broken telephone transmission from here to various random lists of trivia is kinda like claiming that English "has a base 12 system" just because eleven and twelve are their own not-morphologically-analyzable lexemes (and ignoring that already thirteen is then 3+10 and not 1+12).
We have good odds for actually reconstructing the number 10, too, in Proto-Uralic (*luka); but it is apparently in some morphological relationship with 'to count' (*lukə-), so leaves open some room for independent parallel developments. There's even a candidate for 20, but it comes in a few variants (ca. #komćV ~ #koćV) and is proposed to be derived from 'man' (*kojə ~ *koj-ma) (i.e. '20 fingers and toes'?).
It is a stronger argument still, I think, for base 10 already in PU that 8 and 9 are very widely in Uralic derived from 2 and 1, i.e. as something like "two short [of base]", "one short [of base]". As the constructions in question tend to be kinda opaque, I suppose someone could try to plead that 8 originally meant instead 6+2 and was reanalyzed as 10-2 afterwards, and then inspired a new formation for 9… but at minimum I'd want some evidence for the actual appearence of the term for 6 in there too then, which is not at all the case. (One long-standing hypothesis of derivation is instead that a handful of them, like Finnish kahdeksan, contain the PIE root for 10.)
Some additional arguments exist for a special status of 60, based on Permic and Mansi, but this still implies 10 as the primary base and there's again no evidence whatsoever even for "secondary-base" expressions like 60+10 or 60×2.
— Kind of besides all this is that I think I have some very tentative reasons to suspect an earlier, maybe pre-Proto-Uralic, base 4 system instead (e.g. Proto-Ugric *ńëla '8' has already been noted to kind of look similar to *neljä '4').