The trouble I ran into was after eating the ghost, the spawn point of the new ghost was too close to my current position and simply sped over me.
With this and all the considerations you've noted above, I'd probably throw an ML solution at it - imagine it takes less coding time.
Coding that from nothing (and understanding it) could take days. Weeks to months if you dont know ml stuff.
In this particular case I think basic Q Tabular Learning could play optimally, no neural net required.
I could be wrong though, throw a full conv at it without temporal difference and see what happens! (probably a kinda okayish score)
This random exploration is less efficient than back propagation, but at least you don't have to muck around with temporal differences. (And it would be hard to figure out how to tweak the weights for backpropagation in this multi-step game; without falling back to reinforcement learning.)
The problem seems simple enough that evolution via mutation and artificial selection has a decent chance to get good in a reasonable amount of computing time.
But both approaches would be an interesting exercise in their respective domain!
Come to think of it, it might even be an interesting toy problem for a SAT solver! Overkill in the same way as an ML approach, of course.