So it's better to have a sea of properties that are created no matter wether you need them or not? Which would interfere with properties that developers created because they are actually required for their code? Which need rules for all kinds of edge cases right in the compiler?
Don't understand me wrong, I already wrote that it's fine for me if people use the the techniques outlined in this thread - I just don't think it's good to put all of this into the compiler.
Taking you example of FSM: Do you have numbers how many Swift projects use an explicit state machine, how many of those don't use enums to model their state at all, and how big the number of project that use enums for error handling is in comparison?
Besides that:
If someone is really serious about FSM, the first thing he creates is probably a tool for visualization, or even a custom DSL - so creating code for the states is only a small extra step.