Unfortunately, when attacking a subject with a confounder as severe as Bucholz' mechanics issue, you cannot go by sample size alone.
It's not subjective to allocate part (or most) of the blame of his subpar 2008 season to the alteration of his mechanics by the FO, but according to every peripheral available, Bucholz was a much better pitcher than Joba Chamberlain last year,regardless of total sample size and coming off two straight seasons where his overall numbers declined significantly, i believe any neutral source would rather fill out its rotation with the pitcher who's trending up rather than the one who's trending down.
Also, you're trying to use ERA as the overall justification of why Chamberlain is "better" this is incorrect, taking this into account, allow me to ask the following question:
Cole Hamels has a much better lifetime ERA than Matt Garza, but after last season, who would you take to fill out your rotation for one year?
I realize Chamberlain's and Bucholz' sample sizes are not quite as comparable, but it helps illustrate my overall point.