The obvious difference between JPR and Steam surveys, is JPR is telling you what is being sold in a particular small time frame. Steam shows what gamers are currently using, which may span many years worth of ownership and doesn't track GPU's being used for other purposes.
I really doubt the opt in and opt out aspect of Steam plays a significant role in results. It just shrinks the data a little.
Opt-in vs unbiased sampling is a very large determinant in the quality of a survey. This is well established in data science. http://blogs.abcnews.com/thenumbers/2009/09/study-finds-trouble-for-internet-surveys.html. It's due to selection bias - people who opt in may feel they have something to say and that influences their decision to opt-in in the first place.
Again - fine for its designed purpose (a general overview of Steam user's hardware for low-precision decisionmaking). Not at all fine when people are debating tenths of a percent.