Some people view solar panels, whether on a roof or as part of a ground mounted solar system, as a blight on the landscape, rather as they did big ugly C-band satellite dishes a few decades ago. Others see the benefit of solar generation, but see the maintenance of the system as a risk and inconvenience. But it appears that these are in the minority, and that there might be a larger cost related to lenders. In the recent study "Selling Into The Sun" carried out in the US it was noted that there might not be enough data yet to allow lenders to estimate values properly when comparing homes with solar and homes without solar. So while a buyer might be willing to pay a premium for the property, it might be difficult to get financing for the premium.
The problem is known as a skewed database. There may be many hundreds of thousands of data points for houses without solar in the sales records, and this is good since we can partition the data into sets, say on a geographic basis, and still have lots of data to represent the underlying trends. However with much smaller numbers of solar-equipped homes, partitioning the same data leads to a very small number of data points, and lower numbers leads to a higher degree of uncertainty, with much wider margins.
This effect will diminish over time as the lenders get more data to work with and as more properties are equipped with solar so that it becomes more of a known commodity.
In addition we will have more data points on the property risks associated with solar; at the current time insurance premiums are quite high in relation to the underlying risks of loss and liability. I'm sure this is again due to a lack of history related to the specific risks. In "Insuring Solar Photovoltaics: Challenges and Possible Solutions" produced by the National Renewable Energy Laboratory they mention that "PV developers report that insurance costs comprise 5% to 10% of the total cost of energy from their installations, a significant sum for a capital-intensive technology with no moving parts."