We set a candle down on a table, and next to it, a reflecting curved surface.
The light from the candle flame is of course, dispersing, spreading out and hence the apparent brightness is dropping as the distance (from our eye to the flame) squared.
The image in the mirror however is presumably made up of the parallel light rays, those reflected at the exact angle to reach our eye. It seems the apparent brightness of the reflected image should be independent of distance. (It will, of course, depend on the distance from the candle to the surface, but let's assume that's small)
Thus as we move away, is there not some point beyond which the reflection appears brighter?
Going further, could we not go an infinite distance away and the reflection still appear bright?
I'm very interested to hear the practical limitations here (such as the attainable quality of a reflective surface) but also whether the pure thought experiment holds up.