I agree. A lot of the custom masks fit close to the bars, that usually comes from user fit preference etc. not necessarily the mask build. Some goalies years ago wore their mask so close to the bars I can't imagine them getting hit in the cage period, kinda scary. (manny fernandez comes to mind) That being said, I feel as though the center bar is the backbone of the cage. The cage should be a crumble zone, but a strong backbone makes sense to me... again, this is completely speculative, I'm not a mechanical engineer.
Also, the thing that I never understood regarding goalie masks was the interior padding. It's true across multiple disciplines, that you want to have the most snug fit without any voids between the helmet padding and the user's head... but we're seeing things like MIPS and other technologies elsewhere that haven't really been adapted, if at all possible. CCM has "fluid inside" (like the old maltese gel?) type padding I think? I would love to see how that is vs. normal rubatex.
The study is interesting but really there's not a ton of context to it. Like someone said earlier, what's acceptable?