It seems to me that it's very much a matter of what you define as a 'problem'. The basic issue is that the two joining rails are placed very close together, with the result that any pressure that tends to knock the halves out of alignment is turned via the leverage created by this short distance into a pretty heavy force that tries to slide one or both rails along its slot against the grip obtained by the screws.
That grip cannot be great, because there are definite limits to how much you can tighten the screws without damaging the aluminium rails. It might work better if there was some sort of a detent hole where the screws bear, but then those holes would have to be very precisely placed and/or the end faces very square or they would simply act to pull the rails out of alignment. Drilling and tapping for some extra holes would probably help too.
That's not to say that it's too hard to get the system to work perfectly well, and the zero clearance and backer-like function of the rail means that it delivers beautifully clean cuts too.
It does tend to mean though that careful handling of the assembly is needed, especially if it's rotated so that the face of the guides is vertical, with the joining rails to the bottom. In that situation it doesn't take much of a bump for the screws to slip and the rails to move out of alignment.
The guy in this video in demonstrating the system precisely explains what's needed to avoid problems - you basically need to (a) ideally work on a flat table so that the joint is not stressed as above, and (b) to frequently check the alignment with a straight edge:
http://www.youtube.com/watch?v=pVFuP...eature=related
That's fine if the sort of work you are doing only occasionally requires long rips. If I was going to set up along the lines that this guy has to use the system for frequent ripping then I'd want to go with the one piece 3m or whatever rail it is that he has too - checking alignment would wear thin pretty quickly:
http://www.youtube.com/watch?v=NiAr7...endscreen&NR=1
ian