Diablobb81 has done that above and the two year sample size should even things out even further. The only thing it won't do is illustrate the improvement. Your mission, should you chosoe to accept it, is to put together a graph of their qualifying time difference, maybe expressed as a percentage of pole time, and see whether the lines converge over each year or the total period. That should measure the improvment.
You'd do them separately so that you can isolate data from 2010 & 2011. Personally, I thought the gap this year in Q is bigger than last year. Could that be explained by the car characteristics, or DRS/KERS issues?
Possibly..... Possibly not. But I think it was a lot more straight-forward in 2010, there certainly wasn't a quote from Ross Brawn saying the car can really only be optimized for the Q or R, but not really both.
Regardless, I think the argument is not only centered around Q pace is it? JB is slower than LH in 1 lap pace, (by common consent regardless of what the LH huggers would have you believe) but his race pace & more importantly race 'brain' has shown to be at least up there & at times far superior this season to LH. Are we expected to have the same discussion there in due course? That will be fun....but not for the MODS!
There's simply too many intangibles for a definitive result to this I believe, besides I think there's the basics of the agenda being MS is simply too slow to be in F1, that's got little evidence from 2011 imo.