[bojan]

This (binoviewer) may work in principle, but I think it is not worth the trouble (collimation will be the problem... you have to have it accurate down to a pixel size, 10 or so microns).
Just doubling the exposure time will produce the same effect as having double the light collecting area.
And, you are talking about composite video.. which means 25 frames per second (or 50 half-frames per second). That means the exposure time can not be longer than 20ms, so only bright objects - planets Moon, Sun will be accesible, but quite comfortably so.
BTW, BW camera will be more sensitive (due to lack of on-chip filters).
Just remove IR filter (most likely it is there) and you may be surpised with the increase in sensitivity (because of including IR part of the spectrum).
It may be better to invest in exploring possibilities to modify it for long exposures, to go into deep sky photography.

As I said before, to combine signals from two cameras will be very hard if not impossible (due to lack of access to required signals, to bring both cameras into synchronization). And, you will still need mixer, to add only video signals, and not to upset the rest of the composite videl signal (colour burst and color carrier, those will have to be regenerated for the combined video)
B