here is what I get when reducing or closing the vent:
when placing anything near the vent without directly reducing the size of the hole, I got no impact on the response. and I also checked waterfalls in case of some delayed stuff, nothing worth showing came from it. you'd have to create a second acoustic chamber outside the vent while also pretty much sealing that chamber(beside the vent itself) to start getting another resonance worth something at the eardrum. because otherwise, remember that even if the ear does reflect some frequency more than others(it probably does), that signal now has to go back into the IEM or through the tip before reaching the eardrum and make an audible difference. chances are that such signal would be a good deal lower in amplitude compared to what the direct sound from the driver generates at the eardrum. so just intuitively, I wouldn't bet on that idea for making the noticeable difference you're feeling.
now I use the miniDSP E.A.R.S that has a fake ear so I can just try what you explain instead of pretending to simulate it like a nerd and probably missing something. measured with the cable above the ear or down(while maintaining about the same insertion depth), I get basically the same result twice:
so that doesn't seem to explain why you're getting a difference.
the third hypothesis would simply be that when you use it with the cable over the ear, you don't get the same insertion depth, angle, or the same quality of seal.
now I fly away like a captain.