I haven't heard that Fauvels had a tumble problem but there is something mentioned in "Tailless aircraft in theory and practice" that's interesting. Apparently once in a while the AV-37 would do a singe flip at the top of a loop. There are two tumble criteria that I know of. One is Irv Culver's crotch depth to MAC ratio. Using paper models of varying sweep and tapper Culver found that the models with crotch depth (the distance from the root TE to a line connecting tip TEs) to MAC ratio of less than 2:1 could tumble if the static margin was too short. The other is from some WWII spin research the NACA did with fighters with bombs or external fuel tanks mounted on the wings. They found that the spin axis depends on the mass moment of inertia. Basically after a stall the plane will try to spin around the axis with the highest inertia. Normally in a conventional plane the heaviest axis is the longitudinal axis but when you put a heavy bomb load on the wings the lateral axis gains a lot of inertia, usually not enough to overcome the horizontal stabilizer but sometimes enough to cause a flat spin (especially if the static margin is too short). In the absence of a horizontal stabilizer a wing-heavy airplane is free to spin around the lateral axis (tumble). So the solution from a design standpoint is to keep the wings as light as possible, especially outboard, and distribute the mass in the pod as far apart as practical.