How about this.
You are right in one aspect. If we draft two OTs, then we are more likely to get a good OT than if we only draft one.
However, your argument that 3/4 of failed OTs will make good OGs is laughable. You are suggesting that any tackle is more likely to be a good guard than any guard is. If this was truly the case, then there would be no reason to draft a pure guard at any point in the draft, teams would simply always choose tackles or centers.
The truth is that while all offensive linemen need similar talents, the guard position is not simply a second class tackle. To be elite you need a different set of skills.
So, what if the success rate of OTs moved to OG is closer to 1/5? Then what? Then taking two OTs is an awful idea.
What if the idea of putting a statistical percentage on the success rate is pure garbage, considering that a successful transition from OT to OG can mean any number of things?
And finally, the simple truth of the matter is that we're not dealing with abstract numbers but actual players.
The team has very good reasons to believe that Davis will be a very good RT - better than Bulaga. Everyone has very good reason to believe that Iupati will be a very good if not Elite LG - also better than Bulaga. So they had no reason at any point to draft Bulaga.
Drafting a player and then immediately following the pick with a contingency plan in case they fail is an awful awful idea.