The folks who say that you won't hear a difference should try it sometime. Not to say the M10 doesn't sound good, or even better than the 722, since I haven't tried myself -- but you might be surprised. I have done a number of controlled comps of recorders and A/D stages. It's really pretty easy since it is often very easy for one mic preamp to feed two different recorders or A/D's.
I've done comps many times of various options compared to the Lunatec V3 I owned for 7 years or so. I'd bet I'd have a pretty good track record picking out a V3 recording since I ran it so much and did many listening tests with it. If nothing else, the V3 has a very accurate and realistic reproduction of cymbal decays, I can generally pick out a V3 recording in a controlled comp readily by listening for this alone.
Most recently I did a controlled comp using my PSP2 with dual outputs, one fed to a Tascam 680 recorder and one fed to an Oade-modded R44 recorder. I found hearing the differences between the recordings to be very, very apparent -- subtle, but apparent. It took a long time and a lot of listening to finally judge which one I preferred since both had their good and bad qualities, but strictly to differentiate between the two wasn't very difficult. That said, both the 680 and the Oade R44 sounded very good and most people would be happy with recordings from either recorder.
On the need to sync recorders, the issue is time drift between recorders. Even if both recorders are set to sample at 44,100 samples per second, in reality they will sample at slightly different rates because the oscillators used to develop the clock signal are not completely accurate. As a rough example, if you have two recorders sampling at 44,100 samples per second and the clocks are 99.99% accurate (+/- 0.0001), one could be running on the fast side of that accuracy, and one on the slow side. Even with this level of accuracy, the 2 sources over 2 hours (7,200 seconds or nominally 317,520,000 samples) could vary by 1.4 seconds. That is, the 2 sources could drift from one another by 1400 milliseconds, and a drift of more than 30 or 40 milliseconds or more becomes noticeable as an echo or excessive unpleasant reverb.
