I think here ότι is omitted: εννοησας ότι τον ποιητήν δεοι… ποιείν μύθους…και should ότι be here ? αυτός ουκ η μυθολογικός, δια ταύτα δη ους είχον μύθους, τούτους εποίησα. The ellipse of conjunctions is not mentioned by Smyth, Jelf does mention but not οτι
Ok a comment says that επειδή is ommitted here, but why it and not ότι or επεί or anything else? What a fanciful commentary
Here ὅτι follows the participle ἐννοήσας to introduce a kind of indirect statement. This indirect statement finishes in one of the sections you left out:
… ποεῖν μύθους, ἀλλ’ οὐ λόγους, καὶ …
Here the regular adversative force of ἀλλά balances and closes the indirect statement. Accordingly we don’t repeat ὅτι.
Καί then introduces an additional adverbial phrase in light of the indirect statement. It isn’t part of the indirect statement (look at the verb forms; also no ὅτι) however from the context we can see it must have a concessive force [i.e. “and (after all) I myself was no myth-maker”]. Because of this we can tell there is an ellipsed ἐπειδή.
But we could also consider και as copulative conjunction and imperfect is not usually changed into opt in indirect discourse. But the commentators just thought as you think.
But we could also consider και as copulative conjunction and imperfect is not usually changed into opt in indirect discourse. But the commentators just thought as you think.