A friend who sings in various choirs reported that, when rehearsing in an empty abbey, she noticed the echo at the end of a phrase appeared to be sounding flat compared with the pitch of note sung. The effect disappeared when there was a congregation.
The main nave of this particular abbey (Tewkesbury) has pairs of undecorated round pillars of large diameter at unusually close spacing (because there are smaller 'Roman' arches between them, which give closer spacing than in most gothic-arched churches). This large area of exposed vertical surface gives a much stronger first return than in other churches with spindly, well-separated pillars.
My explanation was that the reflections returning to the choir would come from each pair of pillars in turn, at successively greater distance. This would mean that the effective path length of the echo would be increasing with each succesive return of the sound and this would be the same as if it were from a single reflector moving away from the singers. The reflected 'image' of the sound source would, in effect, be moving and would experience Doppler Shift which lowered the pitch.
When the congregation was present, they would absorb the horizontally-propagating sound waves which are the ones that would reflect directly back to the choir. The unabsorbed sound waves above choir height would be angled upwards and their reflections would be lost in the general reverberation of the building.
Does this seem like a plausible explanation, or is there a better one.? My friend has excellent hearing and is very pitch-sensitive, so I don't doubt that she has identified a real effect.