Hi Michael (and group)
I have received one data-set already from a client who bought a number of Audiomoths to test with a view to getting lots because they're so cheap.
The fact that it's a bare circuit-board and microphone means it's not going to be a very robust field-device, especially without some form of water-proofing. The common method for that, and which my client used in this particular deployment, is to whack the device in a zip-lock plastic bag. This is obviously going to cause signal attenuation, as Julie points out, and will likely create reverb and other noise. Its certainly a cheap way of deploying the devices, but again, not very robust/secure!
Having said that, the seven devices from which I received data did successfully record bat-calls, albeit with a lot of noise (but even Anabats will give you plenty of noise if not setup correctly). In the files that I received, the better calls were mainly from loud, low-frequency bats like C. gouldii, but a couple of the devices also recorded Myotis/Nyctophilus and some weak C. morio/V. vulturnus calls.
Four Anabat SD1/SD2 units were depoyed during the same survey and produced better-quality data; however, my client had already searched/filtered the original dataset and only sent me a subset that they thought may contain Myotis macropus. Consequently, I could not make a comparison of overall data quantity or quaity between the Anabats and Audiomoths.
I'm keen to hear others' experiences with these devices. I reckon there's potential there for low-cost, mass-deployment, but I think the trade-off is going to be low quality and probably low signal-to-noise ratio. One of the applications where they could come into there own is for 'citizen science' projects, especially given the generally poor funding levels these days for local conservation and Landcare groups and the like.
Cheers
GF