Trains cross bridges easily as they're either balanced with each half on each side, or half on and half on ground. And other time's just between these stable points. Furthermore, trains run on tracks which reduces chance of turning acceleration.
Railways work well in the center of bridges. One bridge that violates this principle is the Manhattan Bridge; it was built before we had fancy engineering simulations or much experience with railway suspension bridges at all, and so the lower deck is a pair of railway tracks on either side of a roadway.
This unusual design is very stressful on the bridge and has resulted in expensive rehabilitation programs.
The thing about floating bridges is they don't need to span a waterway. The just need to link the part where getting onto the bridge then the bridge itself move to the exit section.
Floating bridges also don't necessarily need to have fixed points to load and unload. With suitable stabilisation and propulsion, they should be able to move to any number of loading and unloading points. We could have a floating bridge per train, vehicle, or any unit or set of users.
there will be a rail link across a floating bridge in seattle opening in 2025.
it was supposed to open earlier, but the contractor messed up the concrete pour and is fixing it at their own cost. but that doesn't have much to do with the fundamentals of the floating bridge itself.
