Why not combine the two? Use their virtual environment as a frontend for a collaborative colony-building simulation (with our "best knowledge" data on the likely distribution of minerals and such incorporated), everything from mining, refining, production, goods transportation, installation/assembly, etc. People could contribute modules that accomplish tasks, with varying levels of design maturity (everything from stub modules that simply take a given set of inputs and yield a certain set of outputs, to actual nuts-and-bolts level of detail systems with rigid-body physics models and CFD chemistry calculations, all the way to real-world tested systems), along with code controlling how individual systems behave in different circumstances. All components would have defined realistic wear and tear over time / various consumables. The ultimate goal for participants would of course be a setup where every module is highly defined, down to the level of nuts and bolts, and every individual component in them can be manufactured by some other system on the planet, in a manner such that the net throughput is sufficient to produce all of hardware required to keep all systems operational plus enough to keep the associated humans alive and comfortable - while having the net mass that would have to be shipped to Mars as low as possible.
It wouldn't be something your "average gamer" would get involved in, I'm picturing something more for engineering students, active/retired engineers, etc, with some funds set aside for real-world testing of the more mature systems. You could generate interest by making clear that systems developed in the environment that reach a sufficient maturity state (passing real-world testing and showing a valuable service to future colonists) would be slated for actual deployment to Mars when the opportunity presents itself.
Detailed 3d environments aren't really a critical aspect of that for some systems (such as refining). But for others, such as transport, they're a critical part of the picture. Even for things like mining, having a good grasp of the types of environments that particular minerals occur in would be quite important - does X occur in this area on hard to access cliff faces, surrounded by dune fields, deep in craters, etc? How can we get it out of there and get it back to where we need it? How can we position each component so as to minimize transport requirements to all others (since one won't find all mineral deposits in the same location)? Etc.