Despite some others' excitement over Carl Malamud's posting scanned microfiche images of 1880s district court cases, I can't get too worked up about it. (Neither can Antitrust Review.) The download's README includes this quote from Peter W. Martin's recent article "Neutral Citation, Court Web Sites, and Access to Authoritative Case Law": "Anyone searching for, analyzing, and then citing authority relevant to a current matter must work with the old as well as the new." Which is true to some extent, but I wonder if Malamud has usefulness backwards by working chronologically forwards. After all, the most recent precedent of a court is the one that a citizen who wishes to know What Is The Law should be reading, not ancient and long-ago-overruled decisions. Starting from microfiche, moreover, only increases the difficulty of making the documents legible, searchable and (best of all) easily cut-and-pasteable. Thus AltLaw, which seems to be working backwards chronologically on federal appellate rulings, has a great advantage in these respects over Malamud's site (and in breadth of coverage, over Project Posner, though the latter has been revised to include Posner concurrences and dissents). Prof. Martin's own Legal Information Institute is a fantastic resource, and one I've used before turning to Westlaw or Lexis, particularly when those have not been freely accessible.
Indeed, while Malamud dwells on "large law firms," I would think the sensible and cost-efficient method of research, at least for a lawyer whose time bills at a lower rate than the databases' time does, would be to search Google (which will hit LII, FindLaw and pdfs from courts' websites, among other resources) in order to find recent cases that are on topic. Then use Lexis's or Westlaw's free document pulling feature to check on whether these cases are still good law. At the end of the process, already somewhat informed of the relevant law, search the paid database efficiently and make sure that nothing has been missed from the area of law and the jurisdiction that matter for the client. Malamud's database isn't useful at any point of this process, and I don't think it's meant to be -- it's more a showpiece to goad the databases into making the underlying government works freely available.
In his letter to Thomson-West, Malamud claims to be confused about how much Westlaw owns of caselaw, particularly that stemming from the days when West was the only reporter: "a large part of the publication stream is tightly interwoven into the very substance of the operations of our courts, with West serving as the either contractual or de facto sole vendor reporting on behalf of the court." Unless Lexis pays for every old case that its subscribers can read or West doesn't bother to enforce its copyrights, presumably a comparison of a Lexis rendition of a case or statute, compared with the West version, would give some idea of what is proprietary and what is not.