As far as the groups of different applicants go, I'm not sure it is as clear cut as you make it out to be. For 2005, the match year of the study, DOs (68.6% matched) did do better than USIMGs (54.7%) and FMGs (55.6%). But, past US grads did worse (44.3%) and Fifthpathways did comparably (60.4%) but only represent a tiny percentage of applicants. Remember also that DOs only account for about 15% of the overall 'All Others' active NRMP applicant pool, presumably diluting their influence on the overall 'All Other' numbers.
This is easy to test with Chi-square analysis or Fisher's exact test, with the only draw back is multiple comparisons.
Here are the p values.
FMGs vs IMGs (p = 0.5, so no difference)
FMGs vs DO (p < 0.0001, so yes DO did better for sure)
FMGs vs US physicians ( p < 0.0001, so yes they did worse. Interesting)
FMGs vs Fifth Pathway (p = 0.4, so no difference)
but...
What this doesn't take into account is what these people applied to...
If 1000 FMGs applied to Family Medicine and 500 matched... vs 500 DOs applying to Radiology and only 100 matched.... it doesn't mean that FMGs did better... it means that assuming FMGs and DO are equal then it's easier to match into FM than radiology.