Not a big issue, IMHO.

Since we're in the business of being boringly safe, you pick a landing site that has bland terrain and few outcrops (like Gusev). You then have real-time imaging of the surface and avoid areas whose contrast or illumination differs from the mean, and optionally add a scanning lidar or radar to image the topography directly and select a smooth flat spot.

If you design well, you can re-use those same instruments for surface science by mounting them on an arm from the rover that looks down. You also have wheels on the end of the landing legs, so they are both legs and suspension. Every time you share use, you save mass...