If I understand correctly, that is, to simplify almost to the point of absurdity.
Observations about the universe don't match our understanding of physics.
DarkMatter handles the problem by adding stuff we can't see to make it match the physics.
MOND handles the problem by changing the physics to match what we can see.
The problem with DM is why can't we see the stuff.
The problem with MOND is why are local physics different than remote physics.
MOND does a better job at predicting observed phenomenon, because it is based on observed behavior. but fails to provide a good underlying reason for the behavior.
MOND does a good job at only some observations, in particular galactic scale ones, but it's original form falls over completely when talking about observations of cosmological scales.
So DM does better at matching observations than MOND does at the largest scales, where ironically the modifications of MOND are most important.
I don't think that's the problem. Data and observations are the same thing. You are saying we need more observations to explain the current observations.
No, we need new theories to explain the current observations.
Observations about the universe don't match our understanding of physics.
DarkMatter handles the problem by adding stuff we can't see to make it match the physics.
MOND handles the problem by changing the physics to match what we can see.
The problem with DM is why can't we see the stuff.
The problem with MOND is why are local physics different than remote physics.
MOND does a better job at predicting observed phenomenon, because it is based on observed behavior. but fails to provide a good underlying reason for the behavior.
Both have edgecases they fail to handle.