The piece of solar roadway in France has already been discussed at length. In short: the are two big flaws in the article:
1) is that the author assumes solar panels on roofs are actually going to happen. WITH the required grid upgrades to handle all the power which people tend to forget about.
What do you mean by "going to happen"? Quite a lot of the existing solar panels are already on roofs. The sensible question is at what level of roof utilisation will solar plateau out? There is only so much roof that faces the right way, has an area big enough to be worth covering in panels, isn't in a conservation area, etc.
The grid infrastructure issue is as true for using road surfaces as for using roof top panels. The very distributed nature of road surface generation means it will almost certainly require numerous grid injection points, similar to the situation with roof top generation.
2) is that the piece of road in France is a test and not a commercial installation so it makes no sense to compare it with a commercial installation
It makes no sense to look at how much a prototype road cost. It does, however, make sense to look at the inherent complexity, which will give an indication of likely volume installation costs, and robustness, which will affect lifetime. I agree that the article does a poor job in this area.
Still let's forgot about the 'solar roadway' in Sandypoint. That is just a gimmick. An author of a serious article about solar roadways wouldn't have mentioned it.
I agree that the Sandy Point setup has no place in a serious discussion about solar roads. What people should include in these articles is things like the German installations of solar panels along Autobahn verges. They look so much saner than paving the driving surface with solar panels.