I'll play devil's advocate for a minute...
If you're really not sure whether you want to do research or clinical work for your career, why not do your MD and part or all of your residency training first. This way, once you're a resident or fellow, you can you can decide on a career based on your interests at that time. That is, once you are finished or nearly finished with your clinical training, you will have a better sense of whether you want a majority clinical or research based career. At that point you understand the clinical world, and and can decide whether you want to go off and be an attending with the money, flexibility, and status that provides. Or, you can decide to go back and be a student, losing all of the money, flexibility, and status, but gaining intellectual stimulation and academic rewards.
Similarly, you can tailor your research to your clinical background and interests after or during residency. This is as opposed to an MD/PhD student who very often will completely switch their focus from their PhD to their residency and fellowship research. Thus, the MD-only research pathway often represents a more efficient training pathway for the motivated student. The student going into the lab, having typically not done any clinical rotations (and very rarely all), has little clue what medical specialty they will enjoy. Thus that student can either not worry about their future career research (a bad thing with the subspecialization of research), do something so broad that it really applies directly to nothing, or take a gamble that the research they do fits with future clinical interests.
Meanwhile, the pre-med really has little sense of what the clinical or research worlds are really about. How can a starting MD/PhD make a good decision about their career 12+ years from when they they finish undergrad? Consider that it's 8 years for MD/PhD then 5 years for residency and fellowship. This probably contributes to the unpublished data (will this ever be published?) I've seen that well over half of all MD/PhDs don't make research a majority of their career. So why spend 4 years of your life training for something you are more likely than not to ever substantially use? Similarly, why should we train so many PhDs who will never use those PhDs?
With medical school payback for research, you shouldn't have to worry about the medical school loans as much, so finances are not so much of an issue. Similarly, one can see in this thread (
http://forums.studentdoctor.net/showthread.php?t=585855) that the PhD really does not gain one much versus other top medical students when it comes time to apply for strong residencies. These top academic residencies are the ones that set you up for faculty positions and future research. Thus, if you are strong clinically (which you must be anyway), research opportunities will present themselves to you in good time if you wish to seek them out. So in the end it's easy to make the conclusion that MD followed by research fellowship is an excellent pathway for research training.
Ok, done playing devil's advocate, flame away