Not sure about the explanation....
Hopefully a chemist can confirm this but...
The organosilane being used (aminopropyl triethoxysilane, APTES) is an agent designed to improve bonding between surfaces which don't really want to bond. Namely the silane group(s) like bonding with hydroxylated surfaces (preferably hydroxylated metals), the amino groups go after organics. So you'd want to have an incredibly thin layer (1 or two molecules thick) to get an improvement in bonding two surfaces, whether glass or glass/plastic, regardless of whether other adhesives are used. Any more than that and the adhesion will be pretty bad - you'll have excess APTES molecules bunging things up.
Secondly, APTES is pretty much clear - the APHA value of <25 is a very, very faint yellow colour, so how much of it would you need to have piled up in one place to get the colours people (myself included) are noticing? A layer trapped between glass wouldn't show up, even if it was thicker than the very thin layer required. The solvent used to apply the stuff (IPA by the looks of things) is clear as well. And as a final argument, if the marks are due to APTES residues (the solvent? the clear solvent?) having not fully evaporated, why does everyone see marks in much the same places? Surely there'd be blobs randomly distributed..? I don't buy it, sorry... I'd like to, but I don't.