I'm not sure if the NASA stuff we see is totally color accurate, but it's the best reference we have, and going by that I think your dust and rock colors are just too saturated. Desaturate a bit and I think you'll gain some realism. Color variation would also help a lot, indeed. But make it subtle! In the first image having a more neutral sun angle will also help both us in judging the look of the scene and you in adjusting it with a more clear basis to differentiate your changes and their effects. This will be especially true for subtle adjustments of color variation for example. Then when you have more achieved a look you think is working with a more "broad daylight" lighting, then you can shift the light to something dramatic like that and see if the texturing holds up and still looks realistic.
All that being said I think that first image looks very promising, especially the dunes and the sand on some of the rocks. What happens to it when you apply it to Mars terrain is not great unfortunately, but that may be looked at as a separate issue, and I would wonder what data that is that you are working with as some of it was quite lumpy. You might try basic MOLA or other lower resolution Mars data just to get some Mars-like shapes (craters, etc.), but an otherwise smooth terrain as your rocks add a lot of detail. Later on you do want to work out the integration of the highest resolution data with your procedural texturing, of course, but for now it may be causing more problems than necessary with your experiments in texturing, and it could be tackled as a separate problem later.
Overall you are definitely making progress! Keep it up, we'll have an awesome animation one day.
- Oshyan
