Firstly i wasn't suggesting you was incompetant, i just gave you the link because i was using javascript rollovers until i came across this and i think this is a much better way of approaching things. That and you said the site was doing had loads of javascript rollovers...

I assume by statistics you mean bandwidth, file size differences? Or do you mean development time? I have never come accross concreate statistics for any of the things i mentioned, this is because bandwidth and file sizes vary from page to page and developer to developer.

Logic dictates that if your file sizes are smaller and more easily cached then it will lead to improvements across the board.

Perhaps someone else will come forwards with some direct comparisons.