Also agreed. And while I don't think it should be possible to break server max speed, I think allowing a rail-docked entity to inherit the velocity of its mothership while undocking should be allowed if only for catapult decks on carriers and other such novelties. To ensure server speed settings are respected we can limit the velocity inheritance to one layer. So if you have a stack like above, the undocked entity would conserve the momentum of the entity it was docked to relative to the entity that THAT entity was docked to. So if you have three rail launchers on a planet surface each going 50m/s with a drone on top, the drone would move 15om/s when docked but only 50m/s when undocked.
Alternatively we can raise rail speeds to the server max setting but then impose a penalty for chain docking, where the top speed of any set of chain-docked rail segments is server max speed divided by the number of chained rails. So in the previous example, if the server max speed was 150m/s, a rail on a ship, station, or planet would move an object with nothing else docked to it at 150, but an object with a second object docked to it would move 75m/s, etc.
We could also implement my previous idea of increasing a ship's effective mass relative to its velocity as a percentage of server max speed using the Lorentz transformation (basically inducing the effects of general relativity with respect to mass using server max speed as a substitute for speed of light), then have a docked entity's effective mass determine whether it suffers a movement penalty for exceeding the rail's mass limit.
There are various ways we can enforce server speed limits on rails, but the point of the matter is they are still too damn slow.
