[plt-scheme] Statistics (V301.5 Speed Up)

[re-adding PLT CC: that was dropped to get the images through]

The two peaks are at 1950 and 2100 (seconds, right?).  And the two
peaks look roughly similar in shape.  So I'd guess there's something
that either happens once or doesn't happen in each run, that takes 150
seconds in total.

As far as beer is concerned: my interpretation of Joe's point was that
we should see a difference in the tails of the distributions: while
the right tail (runs taking more time) should go on arbitrarily (maybe
the machine decides to go off and do a Sudoku while it's running your
benchmark), there should be a point on the left below which we see no
outliers.  What do folks think?