This is all very interesting. The pre-release comment here seemed to be that th H2 was going to be a piece of shit. I am glad you guys did not get burned. Let's see how they hold up in the field.
I think the verdict so far is that they are great when used as what they seem mainly intended for, a point-and-shoot one-piece recorder with a very flexible mic system. Line in type operations are also fine so long as you can externally reduce the input level a bit, due to over-sensitivity of the line input. The mic input is a disappointment, however - and a mystery, as one would have thought that it would use the same input stages as the internal mics, but from the noise level it seems not to. I wish Zoom would come clean and let us know why this is, if only to satisfy curiosity.
While the internal mics seem low enough noise to be useful even for natural acoustic sounds, connecting up external mics seems plagued with excessive low frequency and high switching spike noise as the input noise plots I ran indicate. Mic input -3 dB bandwidth is 50-40,000 Hz. So low frequency is being rolled off maybe too much for some purposes.
Line input -3 dB response is 10 - 42,000 Hz and is most useful with external preamplifier with attenuation network to reduce signal to <-1.9 dBu H2 maximum allowed.
These plots and more H2 technical info to be posted sometime on my site in an upcoming review of this lowest costing feature-loaded deck.