The first step for most of these projects is to make the old data available. Just about all of the old data is on photographic plates, so there are projects at many of the great observatories to scan their archives.
A good example is the project at Smithsonian Astrophysical Observatory (located at Harvard) to digitize the entire plate collection, ranging from ultra-early photographs to sky surveys that Harvard Observatory did during a large part of the 1900s. The plates are scanned, and they are matched with star atlases to determine their exact coordinates and their limiting magnitude and the filters used, if any.
This data -- and lots more from all over, including the epochal Palomar Sky Survey -- is available (or soon will be) using special software such as Microsoft's Worldwide Telescope.
So the old data is being collected and curated and being made available to everyone.
(Note that the old data, being photographic, is limited in how deep they go, so many objects are simply not visible in them, including most minor planets.)
And lots of projects make use of it. The ones I'm familiar with tend to use the data to look for prediscovery images or to extend light curves into the past.
New surveys like the Large Synoptic Telescope will generate cubic shittons of data and are being deliberately designed to separate the collection of data from the use -- it'll collect deep full-sky images every few days and then people will mine that data for things that the people who planned the survey hadn't thought of. It will revolutionize astronomy. Key point: All this old data is sure to be made available in the same way and the very bright people developing ways to find interesting stuff in the LSST will also find good stuff in old photographic plates from the 20s.