Macro tilt-shift? I wonder what kind of blur that would have.
No blur. The point is of course to be able to align the focal plane with the subject. That is the primary use, along with perspective control, of a TS lens.
+1. By using the TILT function and tilting the plane of focus into the subject plane, a deeper DOF is achieved from front to back. This is good news for macro enthusiasts who want to achieve a deeper depth of field without having to resort with the image stacking technique in post processing.
Only if the subject is two dimensional, after all tilting doesn't increase dof, it just moves the plane of focus. For instance the ubiquitous fly's eye macro image would gain nothing from tilting.
I am not saying Canon can't or won't do it, they do love their tilt, and macro lenses and combining them would be yet another "because we can" 17mm TS-E moment, but I highly doubt it. Apart from the above mentioned practical use with three dimensional objects there is the MAJOR issue of the J distance and the tilt angle needed at macro distances.
This all falls into the area of Harold Merklinger and his seminal work "How To Focus The View Camera", Everybody, normally with a "knowledgeable" smile, talks of Scheimpflug and his principle, but the really useful guy for us as photographers is Merklinger and his J Point and Hinge Line. See here for a couple of cool gif's and more info http://www.northlight-images.co.uk/article_pages/using_tilt.html
Basically the longer the focal length the more tilt you need for the same effect, AND, and this is the really important bit for a potential macro tilt lens use, the closer the camera is to the plane of focus the more tilt you need. Bear in mind a 90mm TS-E needs 36º of tilt with a J distance of 6", current Canon T/S lenses have 8º of tilt.