The test I like to apply, and argue should be our standard one, is the same we've generally settled on for videography/cinematography:
The important thing is the intention, not the equipment used. This site is about the art, science, hobby, and business of photography. Photography is the making of pictures by means of a camera. If you are making or doing things that are not photographs using the same equipment (or file types, or software), this just isn't the place for it.
Here, you ask Is photography for technical documentation and/or measurement off topic? On the face of it, photography "for technical documentation" is on-topic: the goal is, after all, to make a photograph; and "for measurement" is off-topic: the goal is to have a measurement.
When it comes to photometry, photogrammetry, computer vision, and other topics, the true answer is often "cameras made for photography are a terrible choice for the thing you want to do, because the precision you want isn't a design consideration". It gets frustrating for everyone involved when that's the answer over and over again.
On the other side, when the question is about technical details of a camera module made for embedded use, the answers are unlikely to be of any interest or relevance to people who are looking to use cameras designed for photography for photography.
Because Stack Exchange tends to be technical in nature, born as it was out of a programming Q&A site, I think we need to be extra-firm in this. There are other places on the network which cater to those engineering challenges. As @dpollitt's answer explains in more depth, keeping this site focused makes it more interesting for and attractive to people interested in the main topic itself.
On the other hand, just as with video, if the question can be put in a frame (heh) which is generally relevant for photography, go for it. In fact, I think this is an excellent exercise: if you find that putting your question in the context of making photographs results in answers that aren't relevant to your use case, your use case probably wasn't on topic to begin with.