recovery web sites are mandated by the recovery.gov keeps track of these sites and currently makes them available in a list of agencies and a list of states (also available as a table). it seems that none of these lists are available in a machine-readable format, though, so the problem of Finding Stimulus Feeds already starts here, because it is impossible to write code keeping track of stimulus sites.
ideally, these lists of recovery sites should be available as feeds, so that subscribing to this feed would allow people to track new sites as they become available. at the very least, these lists should be available in some machine-readable format, so that it is possible to use this information for tracking how well agencies are doing in putting up these sites.
the National Science Foundation (NSF)'s recovery web site does even publish a feed (which may very well be Stimulus Feed #1), but the feed cannot be found by feed autodiscovery, which introduces yet another hurdle for tracking the information made available by agencies. just adding a line like this to the recovery page's head would remove this hurdle:
<link rel="alternate" type="application/atom+xml" title="Atom Feed for Weekly Updates" href="/recovery/weekly.atom" />
it is really exciting to be able to watch this ecosystem grow in real-time. on the other hand, this quick growth (which is only just beginning) also demonstrates that the sheer magnitude of the recovery act will very soon make it impossible to find and use any of this information reliably, if more detailed guidelines are not put in please pretty soon.