By all reputable accounts, SCO has no real case (although there may be some valid issues in their contract dispute with IBM (however unlikely, since IBM can show documentation that they own all derivative works)).
First, if there were actually stolen SCO code in the Linux kernel, SCO has legal issues that they have failed to conform to. First, they can only seek relief from the party that placed that code in the kernel, and not against the users. Second, the courts usually hold that the copyright holder must make good faith efforts to mitigate the infringment from the date of discovery, i.e., once the infringement is discovered, they should notify the party of the details of the infringement in order to give them the opportunity to remove the infringement. It is not a legal requirement to do so, but it does weigh against the concept of damages, for if the infringers aren't given the opportunity to mitigate, then it is generally held that there is little or no justification for damages (the rationale being that any infringement after the discovery where no notification occurs is not damage if the owner is aware of the infringement, but does not mitigate, and if there is no damage after notification, it diminishes the idea of damages befor it).
Second, such evidence that has been presented by SCO has been duly ripped apart by just about everyone. In all cases, the purported code existed in other venues legally available to Linux (and others) apart from SCO. There does seem to be an issue of some BSD licensed code having the advertisement clause removed, but even that evidence points to parties other than direct contributors to the Linux kernel.
Third, the history of the source tree owned by SCO includes the burden of the res judatica applied to the codebase in the 92-93 court battle between AT&T and the University of California (et.al.). Basically, the result of that battle was that all but three files of source code were declared to be freely redistributable. The three files were removed from the BSD code, however, the body of the other files were still present in the Novell/USL/AT&T codebase. This means that any code in the unix code owned by SCO that can be traced back to this settlement is legitimately free for use, and cannot be constrained by SCO.
OSI has a position paper here, http://www.opensource.org/sco-vs-ibm.html#id2791689, that provides a good deal of the history of unix, and the lack of any validity to the current claims by SCO.
[Message Edited]