We have both cheap and high efficiency cells. We want to push the whole cost-performance curve so that the cheap cells get cheaper, the efficient cells get more efficient, and the cells in the sweet spot for any particular use get both cheaper and more efficient.

We also need new tech like flexible solar or window integrated PV to mature, and new financial ideas like selling battery integrated solar in africa like a cell phone contract to be developed.

We're past the stage of crazy inventors in the lab being the only player, now we need salesmen, marketers, engineers, manufacturers, tradesmen, architects, accountants and regulators to actually roll out this tech as widely and effectively as we can.

I've read the argument that it's the opposite: since there is plenty of area left for solar panels (roofs, for one thing), efficiency (in terms of energy per area) is less important than price (energy per dollar).

OTOH, I have also read that for many installations, especially home-scale ones, the cost of the solar panels itself is in the same magnitude as all the other installation costs; in that case, the best way reduce price (and increase energy per dollar) would be to increase efficiency, effectively installing "more energy" at the same time for the same installation cost.

So maybe we don't need cheaper panels, but do need cheaper installations, and a way to do that is to have more efficient panels.

There's also transparent cells, a few companies are trying to market them right now. Lot of areas to reuse, probably low installation fees too.

Interesting! But for every bit of light that goes through, that's light not being used by the panel. Do these transparent cells do any better in terms of effeciency than just using a more effecient, nontransparent, smaller cell combined with a regular window? For example, what gives more energy, a transparent cell of 1 square meter area of 50% opacity, or a regular cell of 2 square meters?

Detail: it's not a "visible lightwave" cell, it captures near infra red and other part of the non visible spectrum, so the usual light energy is not absorbed, it's 95%+ transparent for us, but it yields some voltage. I don't have the figures (I should have looked at); the first company I've seen was Ubiquitous Energy (from a Stanford or MIT lab prof. circa 2013). I've read on some research paper that there are other in the competition.

Maybe the power output is too low.. considering how many zillions of office windows could be converted I assumed they'd be as famous as tesla.

Or you make the solar panels easier to install.

Also have a look at http://www.nature.com/articles/srep18721. It raises the question if it is a good idea to have solar cells which partly consist of lead (Pb).

There might be many new applications for low yield, low cost, lightweight solar panels. How about vehicle external surfaces?

Why should we not hope to get both more efficient panels and for other applications, cheaper and lighter ones?

What's the point of low-yield panels on vehicles?

Depends on how low the yield and how low the price, but one reason for putting small panels on cars today is to run a fan when a car is left in the sun, so that the AC has an easier time cooling the car when the driver returns, saving battery range. Can help to keep the traditional 12v car battery topped up too.

I would argue that working on reducing the cost of installation is the big gains to be had right now.