Hi FireAce51. Welcome to the boards!
I think it simply comes down to how accurate the sample (which you mentioned is short) actually is. In my opinion it's only a few thousand hands and not enough to draw anything more than an entertainment-value conclusion from either way.
After doing some searching I found this 2009 thread on the topic: Typical Win Rate for Micros...
You'll notice immediately that 33BB/100 hands winrate is not sustainable, even at the micro levels.
My only recommendation would be to continue playing within your bankroll and do a results-based evaluation once you have more hands tracked.
Some of the more experienced cash game players on this forum will be able to provide better insight on the differences between $2NL, $5NL, and $10NL.
Best of luck to you!
Reply With Quote