Can't explain it either. I'm not a reg tech.
The swivel o-ring in the LP fitting is essentially identical to the o-ring in a barrel type isolator manifold. It sits in a groove in the center of the LP fitting (the "isolator" portion of the connection) and rides inside the barrel of the inlet fitting on the second stage (essentially the "manifold" section of the connection). So the o-ring will continue to seal effectively even if the fitting is loose, provided it is not so loose that the o-ring is backed off to the end of the inlet barrel where the o-ring can then extrude.
Two possibilities for allowing the fitting to cause a freeflow through the second stage come to mind.
Dacor used to have a nifty feature in their second stages where the inlet end of the LP hose served as the adjustment tool for the orifice. It made adjustment a snap for the tech and did nt require an inline adjustment tool to be used on the LP hose. In practice you pressed forward against it under spring pressure and then rotated it to adjust the orifice, which control the orifice to seat engagement. In order to do that, you have to remove and "e" clip. from a groove to allow it to move forward.
The downside was that if the clip went missing, then in some circumstances the reg could be accidentally mis-adjusted. Similarly, if one of those hoses found its way onto another brand of reg, the difference in orifice depths might create an identical situation where loosening or tightening the fitting might adjust the orifice.
A similar condition can exist with some older LP hoses where the support that holds the swivel o-ring extends past the end of the LP fitting. In some cases those can contact the orifice and move it under friction, creating the same mis-adjustment issue.
Other than those possibilities, I have no idea how the snugness of the LP hose to inlet fitting would impact the potential for the second stage to freeflow (as opposed to just leak at the fitting due to an extruded o-ring.)
