USB2 vs USB3

as far as recording bandwidth tlw is corrent.  it's doubtful anyone will ever need USB3 just because the bandwidth is too low for USB2.  
 
doing the math (assuming recording and output at 24 bit / 44.1kHz sampling rate - (also assuming true 24bit recording, so zero noise floor from the hardware)
24bits X 44100samples/sec = 1.0584Mbits/sec for 1 channel input only
you need an overhead of 2 channels of data stream for output (stereo output), so the output channels need 2.1168Mbits/sec for output.  unless you run more than 1 stereo pair of output simultaneously, this number won't change.
 
USB2 bandwidth = 480Mbits/sec
output = 2.1168Mbits/sec
total bandwidth - output = 477.88Mbit/sec
 
that's the number we are left with to calculate how many channels of input we can have.
477.88Mbit/sec divided by 1.0584Mbit/sec/channel = 451.5 simultaneous channels you can record.
 
this is theoretical, of course.  but even if you consider half of that saying your system (and by system I mean the complete path in and out of the soundcard and computer) is only 50% efficient, you'd still get over 200 channels you could record simultaneously on USB2.
 
I think most audio hardware companies consider the "safe limit" which they will not go above to be 127 channels on USB2 and 72 on Firewire 400.
 
THIS is the main reason we've not seen many USB3 soundcards.  yes, there are some available, but the bandwidth of USB3 is complete overkill for recording.  they're only offering the USB3 connections because of consumerism ("bigger is always better!") and/or convenience.
 
 

Beagle
as far as recording bandwidth tlw is corrent.  it's doubtful anyone will ever need USB3 just because the bandwidth is too low for USB2.  
 
doing the math (assuming recording and output at 24 bit / 44.1kHz sampling rate - (also assuming true 24bit recording, so zero noise floor from the hardware)
24bits X 44100samples/sec = 1.0584Mbits/sec for 1 channel input only
you need an overhead of 2 channels of data stream for output (stereo output), so the output channels need 2.1168Mbits/sec for output.  unless you run more than 1 stereo pair of output simultaneously, this number won't change.
 
USB2 bandwidth = 480Mbits/sec
output = 2.1168Mbits/sec
total bandwidth - output = 477.88Mbit/sec
 
that's the number we are left with to calculate how many channels of input we can have.
477.88Mbit/sec divided by 1.0584Mbit/sec/channel = 451.5 simultaneous channels you can record.
 
this is theoretical, of course.  but even if you consider half of that saying your system (and by system I mean the complete path in and out of the soundcard and computer) is only 50% efficient, you'd still get over 200 channels you could record simultaneously on USB2.
 
I think most audio hardware companies consider the "safe limit" which they will not go above to be 127 channels on USB2 and 72 on Firewire 400.
 
THIS is the main reason we've not seen many USB3 soundcards.  yes, there are some available, but the bandwidth of USB3 is complete overkill for recording.  they're only offering the USB3 connections because of consumerism ("bigger is always better!") and/or convenience.
 

 
The reason to create USB 3.0 compatible interfaces isn't bandwith.  It's mainly to deal with the fact that after Ivy Bridge, Intel stopped supporting some of the backward compatible specs of USB 2.0, and never fixed the problem.  This caused a lot of problems with people using USB 2.0 interfaces in USB 3.0 ports.  Especially interfaces like the Virtual Studio Live by PreSonus and the Roland Octacapture that supported real time DSP effects that borrow the computers CPU cycles to process.  I went to having a $500 paper weight on my 1818VSL when I had to buy a new computer that only had USB 3.0 ports.
 
When I say the industry needs to support USB 3 or USB-C/3.1, they need to support drivers for USB hardware, and making sure those devices work.  These companies have little resource and development money and a lot of it was wasted by being jerked around by Apple on TB.