I think they just went ahead too fast, without enough in-house and beta testing. Possibly they were driven by excitement, and lack of experience in product testing and production.
They might also be worried about competition. Is there some similarity to the Netatmo products? Netatmo already has indoor and outdoor equipment, including wind speed, but so far no sky camera. I have no experience with the Netatmo products, other than having looked through the Netatmo website. They look interesting.
My worry is also related to the apparent lack of concern or lower priority for data integrity. It might be they have put such concerns on the back burner as fixable later.
Of possibly more concern, could be that in a near theoretical view of "crowd sourced data", the accuracy of individual units simply does not matter. In that view of instrumentation, for many tens to thousands of units in a given area, remarkably accurate "local" data can result with a relatively wide spread in the "accuracy" of individual units.
If that were the philosophy (not known, just considering possibilities), it would probably run counter to what many of us seem to want, which is the most accurate and well maintained instruments that we can afford (set in the best practical locations we can individually offer), and then feeding the best data we can generate into a network of amateur sourced data (e.g. CWOP and others).
It is also fun and interesting as a hobbyist or amateur science effort, to have new and different instruments to look at and to work with. However, if they are terribly inaccurate as individual instruments, with no way to calibrate an individual unit, they might be of less interest, or just plain frustrating.