Yeah, its def not quite ready for the large scale, I'll give you that. They're still working on getting the top glass surface figured out... its a complicated piece of glass to get proper traction, strength, and transparency for the PV. They have some people working on it who say its possible, just will take some work/time/money.
Everything else actually works out pretty well on the cost scale, if you start to think about the cost of asphalt roads these days. Then factor in the utility line upgrades that have to be done in most of the country anyway (all the power lines around here don't have insulation on them anymore... they're in desperate need of replacement), the reduction in snow removal costs, etc, the numbers start to make more sense.
I'm not EE, so I can't intelligently comment on the conducting capacities and such. Their concept is that the roads are not primarily a transmittal method, but a localized production method. Therefore they wouldn't be transmitting power over particularly long distances and the loads on them would be relatively low compared to today's setup. I think realistically, you'd have to still have some auxiliary power production for major metropolitan areas, but this could make a huge dent in the need for centralized production plants and massive transmission stations to power all the far out areas. Those areas should be serviced just fine from the roads alone.
Its pretty fascinating... I'd encourage you to keep looking into it, and I'd like to hear more of your input. The guy doing it is an engineer (electrical, i believe), and has thought through all the transmission issues and such pretty extensively. I don't claim to understand it all, but it seems like he has it pretty well figured out. He's also got a lot of cost justification stuff on there to try to show that it wouldn't be cost prohibitive. I think a lot of it is pretty reasonable, some of it might be a little optimistic, but if any of his numbers are even remotely close, the cost factor isn't that big of an issue.
I think the biggest issues would come in the infrastructure changes... houses required to change to DC, or adapt DC to AC to be able to tie in, if you're wirelessly powering cars, the equipment necessary for that, etc. It is all doable, but not quick or easy.