Kajan of Swedish D3 site SvenskaDiablo let us know that a user did extensive testing of the Gold Find modifier in the Diablo 3 Beta, and even made a movie (in English) to document his results.
It’s a nice presentation, but I have to cut to the chase. It turns out that Gold Find works… brace yourself for a shock… exactly as the % indicates that it would! I know, after years of D2’s dubious LCS type maths, I was expecting a revelation of algebraic inaccuracy. But from his testing, 215% gold find does indeed triple (and a little more) the gold you find from monsters, chests, urns, etc. (I’ve never tested it that carefully in the beta, but at one point I played with around 90% GF, and it did seem like I was finding about double the normal gold.)
Of much more interest, to me at least, would be similarly-rigorous testing of Magic Find. It would be much harder to do, since you can’t get a huge amount of Magic Find on the low level beta gear, and you can’t get consistent item drops the way you can gold. Plus, as items drop much less frequently than gold, it would take many more trials to get a sample size that wouldn’t be prejudiced by one especially lucky or unlucky game. But as the beta continues, I’m sure someone will delve into that.