For the test I have in mind, I think you could do it with one of the bigger sounding rockets or a smaller booster (e.g. Falcon I or V) on a sounding rocket flight profile. It'd still cost millions, though. However, I don't see any way to validate a Mars reentry heat shield design or aerocapture without some form of pile driver test. A subscale model could be developed in a couple years or so and would go a long way towards validating the design, and it'd be a lot cheaper than flying a full up CEV mission.

I've seen some of those test articles you're talking about. Pretty slick stuff. I don't know how much of that technology is still in the black, but it wouldn't surprise me if a lot of it is still highly classified. Any country, including India and Brazile, that can launch a satellite into orbit can build an ICBM. There are a lot of countries building missile technology and there's no need for us to make it easier for them to accomplish some of the harder parts.