Personally I think several of the books mentioned could be repaired with errata (robots for example isn't a problem book because of what it lacks content wise, it causes problems because of unclear rules and and issues with game mechanics). The CSC might be a good candidate for a second edition because it lacks a substantial piece of content (it has a lot of weaponry but is comparitively light on most other equipment), in addition to rules mechanics problems (most notably price and function discrepancies between books).
Here is what I think: Like most of the posters I think CSC and High Guard are the most deserving candidates for a second edition. Robots needs heavy errata in the initial chapter (the robot design method is ok but there are so many ambiguities and mechanics problems in the individual sections, and very high prices for very rudimentary robots). Cybernetics could use an errata review of the implant prices, but I don't find it as bad as some say. As far as mercenary goes I think it is actually ok as is, I would rather have content on a new subject than another edition of book 1, it doesn't have major game mechanics problems, and while you could argue there is content you could add to it, I don't think it would be worth producing a new edition at the expense of completely new releases.
On a related note, there really has to be some better review before these books go to the consumer. One reason I say this is because we have books covering the same content 5 different ways (for example Agent, Scoundrel, Robots, and Cybernetics all have their own computer hacking rules that differ from the core and each other somewhat, and a few of these books have multiple variants or unique skill in the case of scoundrel). Also a lot of creation systems wind up having pretty glaring flaws (ie no size or speed for animal creation in animal encounters, robot creation in about a dozen places) where it doesn't seem like they were tested out very completely.
