Also I wonder how the soil will evolve under solar panels, with less light hitting it. Probably better than when farmed, but worse than leaving it alone.
It depends on the native ecosystem, but not as bad as you might fear. In places with dense grasses (some prairie grasses?) or heavy forests, the topsoil doesn’t get much light.
(Deforestation has encouraged the development of algal blooms and other fungal/microbial growth in streams that were previously sheltered from light, contributing to ecosystem failure and other environmental issues.)
To me it is obvious that the soil under solar panels will be extremely healthy and useful, which is evidenced by the abundance of growth underneath all the solar farms in Minnesota fields. But perhaps you think the answer is the other way? Where have you seen only mud under solar panels? How did they build them such that nothing grows? And mud is far far better than current monocrop farming.
Usually the metal profiles are just rammed into the ground, if the ground is too loose they are screwed in and gabions are used if they just need some weight. If none of those are possible then concrete is used.