Let's say most of the total cost of a PV array are from installation with option A. Let's say that option B shifts the balance towards panel price because it's easier to install, but the panels are more expensive. Option B is better than option A if, for the same price, more capacity can be installed. Therefore, if option B increases the price of the panels by more than it reduces the price of installation compared to A, it is worse. Let A have a levelized, $/kw installation price of $20 and a panel price of $80. If B is twice as easy to install than A, for instance, it requires half as many panels to be installed, then as long as the price of the panels increase by less than $10, or ~12% of the price of the panels in A, it is the preferable option.

Here's the problem: even if the panels installed themselves, you can only save a maximum of 20% over option A without also making the panels cheaper on a per-watt basis. Unless panels get obscenely expensive to install compared to the price of each panel, it is very difficult to decrease cost by increasing the efficiency of the panel, because high efficiency, tandem cells such as the one in this article cannot be manufactured more cheaply than existing silicon cells. CVD is obscenely difficult and slow compared to mono-silicon production.