Well I would be the first to admit I am not an engineer, I'm just a lowly fabricator. However, when coming up with our designs our engineer (Aaron) does all of his wiz-bang stress and failure analysis. That said, I'm fairly confident that a boxed frame is always stronger than an open channel and while you are right that the forces generated may be insignificant enough to render the hole moot. In my opinion, I wouldn't want a gaping hole in my subframe connector where the part would be experiencing the most twist.