There is nothing that you can do as a MD/MSci that you can't do as simply a MD. In fact, you can get NIH funding as an undergraduate if you apply through the right "diversity" programs. You can be a part of clinical trials as just an MD.
And yeah, a RO1 is just the top NIH research grant available, but there are other grants such as RO3s and whatever that provide funding. I'm not sure exactly what category clinical trial funding falls under, but they are undoubtedly difficult to gain and obtaining a MSci would be just a small step in obtaining the experience necessary to successfully implement these programs.
I'm not saying that training for a year or so in research is a bad idea, as any experience in the field is a huge benefit. For residency purposes, let me clarify by saying having a MSci may be better than not having a MSci, but as others before me have put it, having a MPH and wanting to work in epidemiology is better than having a MSci and never doing basic science research again. The merits of the graduate degree will be weighed down by directors in their specific program and some will be more receptive than others. For the top residencies, of course, generally the best advice is to do well in your steps and rotations and go from there.
For medical school purposes, MA and MS degrees are near the bottom of what many applicants need for admission purposes. If your grades are bad, do a post-bac. If your grades are really bad, do an SMP. In both cases, the MCAT must be savaged.