New applied technology often starts with the top end and works its way to the mainstream, which reflects the current crop of digital mics. Yet what I'm hopeing to see is a different emphasis, with development focusing on operational simplification and cost reduction for mid-range users. Perhaps sacrificing some of the potential gain in ultimate sound quality for the top dog users, but probably not sacrificing any 'real world' quality for most users while making usage simple and much of the gear commodity computer type stuff.
We have cheap digital USB mics at the bottom end, and we have no compromize digital solutions like those discussed in this thread. But I'm talking about targeting the vast market of mid range gear users.
Run some cat-5 to the mics, plug that into a standard ethernet switch along with a bit-bucket. Substitute wireless LANs for cat-5 if you want. Control it with smart phone apps, a tablet, computer, dedicated recorder or whatever you like. Multichanel as simple as substituting a bigger switch. Same for running a snake- its a simple star topograpy network. Fix the dynamic range of the mics at something reasonable that encompases 90% of what most people record with reasonable quality gear and optimize the useful bits of the 24bit ADC to take best advantage of that. Buffer the output of the mic ADCs with memory and SRC as required. Throw away the unused bits and don't worry about it (we already throw them away anyway, either from the top or the bottom). If that is too simplified or dynamically compromized, put a single high/low gain switch on the mics (or in the app). Users that need to optimize super high or low SPL capability or the absolute ulitmate quallity can stay analog or buy the current digital mics that use more specialized, adjustable and costly gear, but for most users cost goes way down by using commodity digital network gear except the control sofware, probably a specialized network layer and the most important part- the mic/ADCs themselves.
I have unintentionally made recordings of classical material which peaked 30dB below FS and did not suffer from raising the level afterwards. Noise floor of the room was still above that of my analog chain. I'd be more than happy selecting between a single high/low gain range for everything I record and never have to worry about further optimizing levels. Fancy audio gear is cool in itself and we all dig it, but when it comes down to making recordings, I'm more than happy to give that up to simply concentrate on mic placement and arrangement which none of this changes.
This will not satisfy a truely professional user and I don't mean it to. A vast amount of audio gear is marketed to a 'prosumer' level where this applies perfectly. Like many here at TS I'm no pro, although I greatly respect those who are.