We get a fair number of questions that are similar to Email Taken and also not found on Instagram (not picking on any one user in particular, the question just happened to be at my fingertips).
I'm of the opinion that the vast majority of these questions are only answerable by the developers/stakeholders of the particular web application.
In this particular case, it would seem to be "obvious" that Instagram doesn't recycle email addresses, but really the situation could have arisen from some internal policy of data normalization or something as random as a particular IP address trying to log in (not really putting the question up for debate, just to say that it's largely unanswerable).
As such, I propose we retire:
Questions on applications or application features that are no longer available are off-topic for Web Applications as no one will ever be able to make use of the answers again.
which only gets occasional use, in favor of something like
Questions that ask about the internal policies, technologies, and algorithms of web application companies are off-topic because only the company or its developers truly know the answer.
I know that we've used "Primarily Opinion-Based" as the reason for closing these questions, because any answers are essentially guesses, but I think it would be nice to have a concrete close reason for these that we can show to users. To those new to Stack Exchange, saying that a piece of sought-after factual information is a matter of opinion is probably misleading, since it is likely known to a chosen few.
Many of these questions could be marked as "Too Broad" as well, but certainly asking something about data backup policies at Facebook could be answerable in a succinct way if you asked a very small audience of people making decisions at Facebook.
The only downside of having a custom reason would be that it's hard to evaluate what is answerable by the company and what is answerable by an end-user, but something like "Too Broad" is also very subjective at times.
Any thoughts on this? Is the existing close reason valuable enough to keep it for statistical purposes, etc.?
