Wondering why Roland and Bandlab could not partner up again. And bring another version of the VS700 on the market.

Price?

they seem to have hardware provider already, let's hope they move into the controller arena

Several companies make nice controllers (Presonus, Behringer, etc).
 
I want the audio interface separate from the controller (more choice, more flexibility).
ie:  While Roland audio interfaces worked fine, they didn't offer extremely low round-trip latency.
RME and MOTU are both excellent in that regard... as is the Presonus Quantum (RTL is ~1ms at a 32-sample ASIO buffer size/96k) .

So far Cakewalk had own controllers:
* StudioMix (Peavey) was expensive device, without feedback (no LED nor display), rather bad not touch sensitive faders, bad choice of protocol (fixed NRPNs) and bad firmware. Most devices are buggy (self moving faders, wrong direction detection for encoders, etc.)
* Roland keyboards. That was more or less "success". The hardware (in control surface part) has no advantages over much cheaper devices.
* VS-20. Digital hardware processor, fully controllable throw MIDI interface (including gain), compact surface part with feedback (LEDs). But... 2x2 only USB1.1, without dry/wet parallel recording, no standalone mode as hardware processor, no way to control the processor from the device, cheap knobs instead of encoders, rather bad latency, bad drivers, was heavily overprices for what it is.
* VS-100. "All in one" device, with one strip surface and motorized fader, build-in recorder, standalone mode, quite some I/O. But... questionable limitations for recorder and routing, Edirol interface (high latency, far from the best drivers). That still could be an interesting device, for 1/4 - 1/2 of the price they have tried to sell it. But proposed price put it into the place where it could not compete in any characteristic.
* VS-700(C). For that money, it was/is possible to get different class of equipment...
 
All attempts have one common component: instead of compensate the price/characteristics/quality of hardware with brilliant software, Cakewalk has failed in that department.

azslow3
So far Cakewalk had own controllers:
* StudioMix (Peavey) was expensive device, without feedback (no LED nor display), rather bad not touch sensitive faders, bad choice of protocol (fixed NRPNs) and bad firmware. Most devices are buggy (self moving faders, wrong direction detection for encoders, etc.)
* Roland keyboards. That was more or less "success". The hardware (in control surface part) has no advantages over much cheaper devices.
* VS-20. Digital hardware processor, fully controllable throw MIDI interface (including gain), compact surface part with feedback (LEDs). But... 2x2 only USB1.1, without dry/wet parallel recording, no standalone mode as hardware processor, no way to control the processor from the device, cheap knobs instead of encoders, rather bad latency, bad drivers, was heavily overprices for what it is.
* VS-100. "All in one" device, with one strip surface and motorized fader, build-in recorder, standalone mode, quite some I/O. But... questionable limitations for recorder and routing, Edirol interface (high latency, far from the best drivers). That still could be an interesting device, for 1/4 - 1/2 of the price they have tried to sell it. But proposed price put it into the place where it could not compete in any characteristic.
* VS-700(C). For that money, it was/is possible to get different class of equipment...
 
All attempts have one common component: instead of compensate the price/characteristics/quality of hardware with brilliant software, Cakewalk has failed in that department.

no wonder they gave up but at least we convinced them to release the surface sdk to public (i.e., you!)