The MO documentation did (I strongly believe but cannot find at the moment) contain a phrase like: "please ask only questions you actually have" in the sense that MO should be used (mainly) as a tool to get information one needs and not as a means to disseminate interesting information (there might be people that have a different opinion on this, but this is at least one not too rare point of view).
In that sense, no in general one should not ask a question whose answer one knows (before asking). And, I agree the phrase form the documentation encouraging posting question and answers should likely be modified for MO.
Related to the example you give (added to clarify, I mean the abstract one not your specific case I did not look into this in detail as this is a general request for discussion) there is a point in general SE documentation, which was not explicity in MO documentation I think but implictly understood to apply (interpreted properly): Avoid overt self-promotion. See https://mathoverflow.net/help/behavior for details.
Thus also under SE "rules" your specific example might be excluded.
That being said, I can imagine situations where it is still reasonable to ask a question where one knows an answer. In particular, if there are (or at least one suspects) multiple different answers possible. An example could be a list-type question where I would consider it as perfectly alright for OP of question to also contribute their own answer(s). (One might do so also in the question, but it can be clearer to maintain a clear distinction between question and answers.)
Or, if somebody where, say, interested in diffeent ways of proving the same result (while knowing one or some), to hava quetion 'In which ways can one prove this' and also answer 'This is one way.' would seem fine to me, as for most relevant matters it would be the same as 'Q: this is a way to prove this are there others?' no answer.
Also, the example that is mentioned in the blog-post was such that other answers than the one by asker where still possible (and it was sort of likely this was so).
Mainly, I would see an issue with question-answer packages of one person that are more or less finalized from the start on, without much reasonable possibility or even interest from OP of having others' contributions in addition.
To avoid confusion: Of course the situation for learning the answer after having asked the question is completely different and here one certainly should provide the answer, or make clear in another way that one no longer has a need for an answer (say via flagging for closure) but the former seem preferable. But as far as I understood this was not the question here.