Last century when I was designing for this (and other) market(s), I speculated about an active mic input expander for when a mixer just needed one or two more inputs.
I scratched out a design powered from the mic input's phantom voltage. Just imagine one NPN LTP per additional mic input. A variable resistance between the LTP emitters could act as a relative gain trim. The LTPs would be biased up at some fraction of the phantom voltage and collectors would all be summed together into + and - inputs . The input current signals convert back to voltage in the preamp's input resistive termination (approx 2k).
This mic extender box could include polarity swap switches, pad, mute, gain trim, etc. I even speculated about phantom powered mics still working and they probably could but loading down the input DC operating point, while supplying larger signal swing was working in the wrong direction for optimal headroom.
I am pretty confident this "could" work, but never got as far as melting solder, because the business model was absolutely terrible... why sell a customer a $20 accessory gadget, with miniscule profit, when you could just sell them a new larger mixer. ;D Back then the price of new mixers were dropping so the upsell was relatively easy for dealers.
When combining identical microphones doing pretty much the same thing, passive combining (like Gene suggested) can work... for more money than it is worth (IMO) dual primary mic transformers could combine the two mic signals. To keep mic termination impedance correct you would probably give up several dB of signal level.
JR
