@simon_brooke I do not know much about language construction. So, I am just thinking this through via reasoning.
It seems to me that perhaps the only reason that one would want to replicate the constructions and inconsistencies of Common Lisp would be if they expect that a large part of the people that will be using their system come mostly from the Common Lisp world, and that the language creator wishes to make the transition for users of their system to be as seamless as possible for Common Lisp coders.
But if that is not the case, then I think it would make sense to try to create a system so that forms are as predictable and intuitive and consistent in their functioning as possible.
I think that avoiding assorted gotchas for eventual users is what language creators should aim for.
Remembering piecewise behavior (ex., in this situation, it works this way, but in that, however …) is unnecessary mental overhead, IMO.