Would it help to use multiple sensors at different positions and use the common points to filter out the 2nd+ bounces? I think this would work for, say, a mirr

Would it help to use multiple sensors at different positions and use the common points to filter out the 2nd+ bounces? I think this would work for, say, a mirror but what about something like brushed metal? Are those other bounces scattered enough that multiple perspective still produce an error?