As someone with a PhD in Pavement Engineering, and an active researcher into pavement design, let me say this is a classic case of someone thinking that because something looks simple it is. Pavements are the most complex civil engineering structures to design, because they are the only structures designed to fail in fatigue. My wife showed me their video the other day, and all I could do was laugh. Reading their FAQ now, shows they've never asked an actual pavement engineer for their input (and FHWA funding shows nothing, in fact googling shows that they're not even really being funded by the FHWA research budget but by the Small Business Innovation Research (SBIR) program i.e. this is money to promote small business, the research is a secondary goal).
Just a correction for you though - there is not really an AASHTO testing protocol, that was a one off test done 50s and 60s. Now, most proof testing of these types of innovative designs are done by accelerated pavement testing.
Before we even look at the engineering, look at the cost: the highest cost pavement currently are precast concrete slabs, which are similar in some ways to this idea (except they are 50 times the size). They cost about $3 million per lane mile to install. There are over 8 million lane miles of public road in the US, so their idea in their video of covering all the roads in the US would only cost $24 trillion (or nearly twice the US annual GDP) assuming they could get the cost down to that of concrete... Assuming for the moment that the solar panels themselves are cost neutral, just the cost of the glass and support structures would make this impossible to afford.
From an engineering perspective, you have functional and structural criteria. Functional are skid resistance, spray, noise and light reflectivity. The glass would polish, resulting in low skid resistance at high speed, and bad light reflection. Their textured surface would be OK for low speed skid, but really bad for noise and spray, even with drainage between the panels. Many new pavements have a porous top layer for this. Their paving stone like pattern would be really bad for noise (like block paving). Putting LED lights into pressure sensors for animals would be fun, but probably not reliable, and on roads you have to have systems that are reliable because either drivers can trust them, or they are a waste of time.
Structurally, the fact that they refer to gross vehicle mass is a dead giveaway that they don't know the first thing about pavements... The critical number is wheel load. Their panels look to be an awkward size between an interlocking block paver where the wheel load is spread across several blocks, and a concrete slab. The panels would need to be connected in such a way that they can expand and contract, with sufficient load transfer between panels for the entire surface to act as a continuum. With this size of panel there would a lot of flex at the joints, which would break most materials. Concrete slabs get joined using 1 inch dowel bars... Assuming these were placed on existing pavements, maybe they would work, but my guess is that they would get beat up quickly by highway traffic.
Then there is a question of life cycle assessment. Their "numbers" page shows they also know nothing about this either. They just include the benefits... There is no measure of the system, including manufacture, construction, maintenance, etc. They also don't have albedo measurements, etc...
So, to conclude, I don't think this idea is going anywhere fast. Their first step should be to hire a pavement engineer. Then they need to do some lab testing, then use their $1.7 million for an accelerated pavement test to determine if their design can work as a road, before they do any more messing around with electronics... At least their idea is not as silly as the people who want to put piezoelectric generators into pavements to capture all the "wasted" energy...