1)as you have to use amps then they give you DC linear output.
2)light also "echhes" from the wals and this echo is too weak to pay any attention to it. the original sound is much more louder (of course if you intentionally drive with your bot in a room that has a big echo factor.
u cannot sense the 'direction' accurately with only two microphones ( our ears audio imaging is much more complex , don't compare microphone with our ear)
(considering 2D) u need atleast 3 mics , the last microphone is to detect the direction(ie front/back)