MOND does amazingly well at galactic rotation curves, less well at anything else. If you think it started with Vera Rubin in 1966 MOND seems natural, but if you know that it started with Fritz Zwicky in 1933 than dark matter is easier to believe.
MOND only really does well on galactic rotation curves because it has free parameters that are tuned to "predict" the correct answer for galactic rotation curves.
MOND only really does well on galactic rotation curves because it has free parameters that are tuned to "predict" the correct answer for galactic rotation curves.