I have a somewhat different take on floppy pans and two-piece chassis.
Basically, a slot car chassis vibrates due to a number of factors: motor vibration, axle assembly imbalance, roughness of the braid path and track, etc. The body, whether hard plastic or Lexan, also vibrates or resonates, both directly and sympathetically.
All of these various vibrations can interact, either cancelling or reinforcing each other resulting in negative effects on handling. Floppy pans or two piece chassis assist handling by "de-coupling" the chassis vibes from the body vibes, preventing these interactions.
Additionally, de-coupling the body from the chassis assists handling in turns by delaying the transmittal of body movement to the main chassis.
There seems to be a point where the amount of relative movement between the two parts provides maximum benefits. Too much relative movement between body and chassis, or between chassis parts in the case of a two-piece chassis, can have a very negative effect on handling characteristics. Generally, movement amounts above the "sweet spot" result in a very loose or tail-happy car and movement amounts less than the optimum amount result in less slide or looseness than will give best cornering speeds.
It is pretty common now in 1/24 non-winged racing to damp or control relative movement between chassis parts using filement tape strips and the width and placement(s) of the tape can give very noticeable effects.
The choice of body material (and therefore flexibilty) makes this very clear, but material choice also affects weight of the body and thus can confuse the issue. However, on the commercial side we notice handling differences between Lexan bodies and the clear PETG material some companies use for vacuum-formed bodies. These materials have very similar weights yet in some instances the handling differences are large due to the flexibility difference between them.