Mirror of the best candidate showers page.
First I ran all the AIRES showers through using the whole San Gabriel array. There is no difference in the energy distribution.

I modified the likelihood code to loop over sites instead of over shmoos. Results, if anything, are worse.
shower 1