We all know that kid who would learn stuff regardless of what environment he was put in, and his opposite. And then probably a fair few in the middle where the environment actually matters.
When you're looking at outcomes, how do you know which school is worth paying for? I'm facing this issue myself right now for my primary school kids.
The big issue statistically is selection. For an analogy, think of cars. Are Volvos safe, or do they just attract safer drivers? It's not impossible to tease out from stats in principle, but has anyone done it? With schools it's even worse. Is one school in a richer area? What about the cost, that will skew selection too? Does one school use an entrance exam? And the really big one, does a school kick out underperformers?
With respect to coding schools you have to wonder whether the "best" students are already creamed off by universities, such that Lambda School and their like are really just salvaging a few people who fell into unfortunate circumstances and are trying to help themselves out. Along with some unfortunates who are never going to be coders but buy the premise that you can be taught these things in a short period of time, you just have to pay.
In any case it seems like they are forced to produce numbers quite fast for business reasons, and this has caused them to present the stats in the most positive light possible. The virus situation is not an entirely crazy excuse, but it will be interesting to see what happens.