Scriblio is a fantastic Wordpress module that makes it easy to build a discovery layer over your regular library catalog (or to make a completely new one). It is completely free and built by folks how care about libraries and users: Casey Bisson, Lichen Rancourt and Jessamyn West.
It seems to me that there are different people hacking away at OPACs--trying to mine & mash them and teach them new tricks. Efforts like Scriblio, LibraryThing for Libraries, John Blyberg's Drupal-based AADL.org, Library Elf, and more.
I think we can reach some sort of synergy if we, at lease, share some code =)
Just counting the different implementations for "getting stuff out of a Millennium WebOpac record", I can count 3 (Scriblio's, AADL.org's and mine)... and they´re all PHP. And I can tell you I didn't use code from any of them (at the time I started I think Scriblio only had MARC import capabilities. Darn!)
With the Berkeley Accord, there might also be some standards, to boot. =)