We definitely need an experiment like the one you described. The potential benefits of aerocapture are enormous. All I'm saying is that for manned missions, the potential risks of a full aerocapture into orbit might be more than we're willing to accept (at least at first). The variability of the martian atmosphere would make for some tricky flight path corrections in real time.

Using aerobraking to perform a partial aerocapture could allow you to burn off a lot of the excess velocity, say 50-75%, before orbital insertion. That would give many of the benefits of aerocapture with less risk. Odds are, even with aerocapture, you'd end up having to do some significant burns to get the orbit right. Whether those burns would require a lot less, about the same, or more propellant than a partial aerocapture-burn remains to be seen. Aerobraking would help achieve the final orbit but you wouldn't want to take months to do it like they did with MGS.