While the evaluation for sky coverage vs perturbation has one data point that decreased with the final revision of the star ids compare function, the real reason is not due the star ids compare function but due to the extra catalog narrowing (removing nearly duplicate stars). Performance improves everywhere but on the very bottom, though, so I think it's ok.
Probably should play around with that duplicate removal number a bit more.
While the evaluation for sky coverage vs perturbation has one data point that decreased with the final revision of the star ids compare function, the real reason is not due the star ids compare function but due to the extra catalog narrowing (removing nearly duplicate stars). Performance improves everywhere but on the very bottom, though, so I think it's ok.
Probably should play around with that duplicate removal number a bit more.