I find that as long as I place the meter in the light path between the camera and the subject, and make sure it is aligned, it doesn't seem to matter so much whether the meter is very close to the subject or half a mile in front of it (given how far that light has already traveled, perhaps that isn't surprising). It's more important that I don't personally get in the way of the incident light on the meter, which often means ducking down and holding the meter up a bit.
I took an OM-2n with a perfectly adjusted internal meter (thank you, Mr Hermanson) out into the snow and sunshine today, despite the -22ºC temperature. I used the external incident meter I mentioned in the first post, as I wanted to know how it would perform in snow and bright sunshine, which fools reflective meters every time unless you remember to compensate and do so correctly. I'll try to develop the film tomorrow if the boss allows, and we'll see.