I have been starting to feel like the development of madmapper's video mapping side has not really been going anywhere in a while. That is kind of a shame since it is such a great software, but maybe garagecube is working on something big?
So just a few thoughts about I would like to see developed on madmapper:
The spacial scanner, this has a lot of potential, like native soft edge blending and video alignment for multiple projectors. Shape detection for automatic creation of quads/shapes. Camera correction filter for live video would be really nice.
Drawing custom shapes (like masks) and being able to warp them.
Bazier warping for masks.
Support/creation of 3D models, to get easier mapping on advanced surfaces. Been thinking about getting Unity pro for this, but still feel it is a little too expensive.
Fading between presets. This has been discussed before, but I think it is time to take it into consideration again since mapping requirements are getting more advanced every day.
More blending modes.
Multiply Syphon Inputs. This has also been discussed before, but I still think it would be great to be able to just choose syphon inputs and get them alinged next to eachother on the input screen. Maybe even make the ability to choose different input screens for different quads.
Adding quartz composer patches to different quads? This could get a little too messy.
Being able to make groups!
And as someone suggested before, making the "input" and "output" mapping go in sync, so you don't have to map everything twice.