The easiest way surely would be to get hold of the Alan Blumlein patent from 1931. The full text is no longer available on the internet to my knowledge but someone must have it?

You will often see references to it in passing, Michael Gerzon mentions it in his article on Stereo Imaging, but Blumlein did calculate that for phantom imaging to work (and this is what a true soundstage is), the 60 degree angle (30 degree to the normal) was optimium.

He also accepted the failings of the system, particularly with regards to low frequency sound localisation, so it would be interesting to see someone post the original patent document.

As for toe in, my understanding was that the speakers should be toed in facing the listener for optimum stereo. The reason we seen to have deviated from that, is that the off axis behavior of many cone drive units can be kinder on the ears than the on axis (or so a manufacturer tell me!). There is no excuse for Naim loudspeaker designs other than domestic acceptance and bass reinforcement/impulse response;-)