Well, I would think that being below what is recommended might, in theory, add some benefit--as in poor (0), good (256), better (1x = 512), better yet (> 1x = > 512 < ?x), ..., and optimal (?x). I really wonder how sound the at least 1x recommendation is. That is, did they really do tests to come up with at least 1x? I really doubt that the difference is 256 on a 512 vs 512 on a 512 would be measureable statistically. |