As a test, starting on a smaller and simpler structure, at the start of this month I made two mirror image walls layered from 80 thou." Evergreen sheet with Slater's Flemish bond 20 thou." brick Plasticard one laminated with cyno and the other with normal (Humbrol) solvent and left them to see what happens. Well, after about two and a half weeks the cyno bonded wall is still dead flat when laid on a mirror while the solvent bonded one just shows a chink of daylight, which wasn't there at the start. Conclusion, I'm planning on laminating with cyno and keeping the solvent for bonding corners and other places where there is direct continuous support (e.g. round a floor or ceiling or at a corner).
Incidentally, if you build compartment stock, don't glue the partitions to the sides, but do ensure they are all the same tickness and in line; I guess building internal walls and intermediate floors would be much the same. If glued, there's a tendency to create a slight dimple effect either side of the partition. I build a NG Third like that just over 30 years ago and the sides are still pretty much flat. They're not that thick either; a layer of 10 thou. clear overlaid with a 10 thou." white panelling fret. Of course, the bodyshell is glued at floor and (false) ceiling height and at the corners. On the other hand, a side I built for a 4mm scale 1930s Collett full brake just three or four years earlier, laminated from three layers and over 1mm thick, has warped all over the place.