It's a good question...but no. There would be no way for the air from one tank to get to the other. At best, the IP drop on inhalation in one first stage would cause enough IP drop in the other first stage and cause it to start flowing air as well so that both first stages may feed the second stage you are breathing from. More likely you would just get some very interesting resonance resulting in both first stage seats getting rapidly cycled and beat to death.hoosier:Can you equalize both tanks through LP ports?
Now, it you set it up so your primary first stage had a high intermediate pressure (145 psi) and your secondary first stage had a low IP 120 psi) it's possible that you could bleed one tank nearly dry an then have the secondary first stage begin supplying air through the connecting hose and primary first stage to the primary second stage.
Of course, in the event of a second stage freeflow, you would have no way to isolate the freeflowing regulator either so you are gaining capacity but not redundancy like you do with independent doubles.
I see the same problem with the universal manifold mentioned above. It will increase your total capacity, but does not provide for a redundant regulator and is no better than the "pigtail" and "cheater bar" systems used in the past in that regard.
Basically the problem is that doubles give you enough gas to get yourself into some serious decompression and if you get there, you need redundnacy to ensure you can get your self back safely. So if you are going to use doubles, use a modern isolator manifold or use independnent doubles, don't use something else that will be less than fully redundant.