In terms of light, a single bottle is insufficient, but ten are enough to light up a mud block-walled, corrugate metal-roofed schoolroom. Since there's virtually zero monitoring of kids' eye health and ability, that can mean the difference between learning to read and learning to look like you're paying attention for young students.
A few things to address:
Night: these things don't produce any light on their own (they're essentially low-cost deck prisms). A lot of people in-country at the time I helped promote these were excited at the possibility that a full moon might produce sufficient light to see, but it's just not enough. You need to supplement these things once the sun's down.
Windows: most of the buildings in rural West Africa are build of mud block, which isn't the most stable or solid of construction materials. It's cheap and readily available, but needs to be built thick and without large windows, or with significant reinforcement to any large windows, or else it crumbles. That means there's very little light coming into these rooms, even in the most sun-kissed of places. Bottle bulbs are an order of magnitude cheaper than installing iron-reinforced windows.
Transparent/translucent roofing: when this stuff is available at all, it's questionable whether it will fit the corrugate pattern of an extant roof, and the quality of the translucent roofing plastics we saw in The Gambia and Senegal was quite low; it would go milky and then opaque over just a year or two, and cost more than replacing a bottle bulb would. By contrast, you can purchase corrugated tin roofing and roofing caulk in just about any West African city, and relatively affordably at that. The bottles themselves can be found cheaply or free anywhere that tourists or aid workers spend time.
Black caps: it's not the color that matters, it's covering the original cap with something more durable. Once you've assured that there's no leaks in the bottle-corrugate metal interface, the cap is the most likely point of failure on these bulbs. The sun is STRONG in these places, which is a lot of stress on opaque plastics; once weakened, a small crack is enough to emit water vapor, and so a bottle bulb's contents would evaporate, requiring replacement.
The real tricks involved are getting a good seal between the bottle and the corrugate tin (we used roof caulking), and making the whole process relatively efficient (it can take a while to punch these out). Once you've got that down, the production and installation process can be taught to locals, who in turn can turn a profit with their labors. I worked with one carpenter for about a year before returning to the states, and he's since picked up a few contracts with the local ministry of education and a few individual customers to install bottle lighting systems.
I would be happy to go on about these things for HOURS (and have, in the past)! They're a niche solution, but they fit their niche extremely well.