I’m always concerned/skeptical when I see a bike that permits a non-insignificant load to be placed behind the rear axle. My rudimentary understanding of bicycle dynamics suggests that in a curve, there is a frame of reference centered on the rear wheel’s contact patch. In a left turn, the front wheel, the frame, and the front half of the rear wheel are moving to the left – in the frame of reference – and the rear half of the rear wheel is moving to the right.

But also, in a left turn, a bike must lean to the left. So net-net, the front wheel and frame are moving leftward and also leaning left, and only half the rear wheel is leaning left but moving right. This mismatch works against the stability of the bike, but it’s small and so is generally ignored.

Putting a load behind the rear axle would exacerbate this mismatch, getting worse for taller or further-back loads – due to longer moment arms – or when heavier (more momentum against the lean). It’s not that less stability makes a bike unrideable, but it does mean oscillations can develop, potentially getting out of hand very quickly. Whereas hunting oscillation for a train carriage is unpleasant for the occupants, oscillations on a bike can throw the occupants off.

Maybe one day someone will develop a yaw damper for bikes, to actively compensate for lean stability at the rear.