There are alot of speed tests on the web that look at javascript libraries like jQuery. All of them have been getting faster and faster. I used a MooTools template, which was actually pretty easy to set up. I pointed it at my blog rather than the typical Shakespeare script. I also included the size of the libraries, which I think is important.
One thing I noticed is that there are ALOT of discrepancies. The number of found items varies greatly. MooTools especially fails on a few common selectors. I was actually pretty liberal in downloading their library. I got everything except the plugins. Not sure what is going on there.
Additionally, I think its much more apparent using a real website how little these speed differences can be noticed by the human eye. Much more importantly should be the accuracy of the findings.
See the test here: http://www.commadot.com/jquery/speedtest/
Whatya think?