I ran a PT3 stat I threw together over some large samples of FR at 50nl, 100nl and 200nl (at least 300k hands for each sample), and here's what I got:

50nl FR - 0.183 VPP/hand
100nl FR - 0.311 VPP/hand
200nl FR - 0.444 VPP/hand

Which seems similar to what we were getting before. Of course my stat could be f'ed, but I don't think it is.

At much smaller samples (at least 70k hands for each level), I have

25nl FR - 0.122 VPP/hand
10nl FR - 0.057 VPP/hand

Good luck guys.