A problem is that a lot of these techniques need a clean background to really see the streaks or shimmers associated with turbulence. In a laboratory setting, this isn't a problem, but an aircraft isn't always pointed straight up or at a clean background. This seems more plausible during takeoff when the nose is more pointed to the sky than cruise or landing, but even during takeoff it seems like a stretch. Also, I doubt that a single star would provide a good enough light source to really distinguish between clear-air turbulence, clouds, or even birds for that matter. It is an interesting idea but unfortunately I do not think it would work in practice.
[1] https://en.wikipedia.org/wiki/Schlieren_photography [2] https://en.wikipedia.org/wiki/Shadowgraph
Firstly, we're talking about clear air turbulence here. So by definition, not in clouds. Indeed most longer flights happen well above typical cloud layers.
Im no aviation expert, but I'd think there are 2 issues in play here;
A) obviously this approach only works at night. And
B) during the cruise phase a flight is flying at constant height. So you'd "look forward" to look for turbulence. But in that direction you're gonna look through a lot of atmosphere before you see stars. Turbulence along any of that path would show, but you can't detect distance. You dont want to be turning on the seat-belt sign 2 hours before the turbulence happens.
It's not a bad idea, but probably not helpful.keep musing on the problem though, it's definitely worth solving.