If you think that it is the management, specifically ZeniMax, that causes the release date problems then we ought to include Fallout: New Vegas as an on time release. After all, Bethesda was publishing it. Their managers were definitely sitting at the table when the release date was announced. We should also exclude Arena and Daggerfall, considering Bethesda wasn't 'aquired' by ZeniMax until 1999.
On Zenimax's formation in 1999, see my edits in my last post and sorry for not being a little clearer on that end in the original post. On New Vegas, I generally try to divide games that are developed internally from games developed externally - the actual developers have significantly more input and the publisher has significantly more involvement when the game is being worked on internally, which means that it's easier for a developer to push for a delay of an unfinished game (something that, given New Vegas's state at release, probably should have happened for that game as well).
Honestly, even if you want to include Morrowind's late release, you can explain that away by saying that the team responsible for Morrowind had never attempt something on that scale before. Similarly, the Oblivion delay can be explained by attempting to rush it so it would be a 360 launch title. Of course, I wouldn't agree that the Oblivion delay was caused by such an attempt.
But the fact that you're searching for ways to "explain away" late releases is specifically the problem I'm having with your post. You shouldn't try to explain them away, you should consider them and then consider things like the scale of the project involved when thinking about just
how much weight they should be given in an anolysis, not try to look for reasons that you can outright exclude inconvenient data. And again, the explanation you're using with Morrowind hurts your stance more than it helps it - if Morrowind
was delayed because they hadn't attempted something on that scale before, then why wouldn't you expect the same with Skyrim, where the current team is making a new engine for their first time (this being the overall studio's first new engine to be developed in-house since the mid-1990s, when creation of that sort of thing would have been wildly different)? That seems to be some rather significantly new ground for them, so it doesn't seem to make much sense at all to claim that Morrowind's delay can be "explained away" because of the scale of that project when the scale of this one is arguably larger in development terms.
EDIT: As for ZeniMax coming in later, remember, the key guy in Bethesda's early history was Christopher Weaver, and he's gone.
He's almost certainly not the only person who shifted from Bethesda to Zenimax in the early days, and I wouldn't be at all surprised if their corporate policy on internally developed games hasn't seen significant changes over that time. Their policy on releases in particular seems to have seen some revision specifically as a result of Daggerfall's release, and that especially is something that I'd expect is still at least partially intact.
EDIT: This is a rather complex sort of thing to be looking at in general. That's why it doesn't make sense to discard
any of the data that we have without seriously considering it. I don't necessarily think that the delays involved with the last two games in the series are going to result in a delay with this one (I think the new engine is going to be more of a factor than anything, though I do expect a delay), but the apparent trend they have for delaying their games is still entirely worth looking at.