@Nicholas_Seward please consider turning the Helios into this infinite printbed printer … btw … any recent developments? I’ve seen the news got blank on Helios lately.
Is SCARA arm the better solution here? The most of its work range is not used and the resonances will be IMHO higher than with Cartesian system. The print quality of the belt systems (like very nice Blackbelt) seems to me not to as precise as top profi 3D printers yet. Unfortunately the material palette is also limited.
@Tomas_Vit Sure they are. But the point here is making a compact infinite printer. For the range - imagine a really tall part. It will require full extension when the front will reach the farther end of the slopped bed.
A HELIOS on wheels would give you about a 300x150xinf build volume from a 250x250 bot. It is a project on my list but I haven’t had any project time this summer. SCARA printers only have compactness and possibly affordability as an advantage. If size and cost aren’t considered then you have to go cartesian.
For remote areas or other areas where linear rails are hard to come by then such printers with low part count and ideally 3d-printable would be the only solution. I Would guess that those remote areas would be candidates for long prints too. And these areas are not known to have the flattest areas to have a wheeled printer run on.
@Florian_Ford I am working on some tech to deal with less than flat surfaces. There might be some artifacts if the surface is really bad but the parts should be functional. I think finding large flat floors is pretty easy almost anywhere.
I was thinking of the need for such printers in disaster relief so they should be quite self sufficient. If there was a hurricane or flood those floors would be pretty much non-existent. In those cases maybe the hang printer or a infinite build area self-sufficient printer would be good helpers for printing fixes, jigs, maybe furniture or other larger, construction-scale parts, maybe even built for concrete printing.
I very much doubt a SCARA arm is of advantage for high speed printing. With enough structure ($), and enough precision ($), and strong enough motors ($) - I am certain you could achieve high speed, at a cost.
A simple CoreXY frame using inexpensive extrusions has structural advantage. Pretty sure you could get equal speed at much less cost.
Of course, I am just guessing. Better informed opinions (based on hard data) are welcome. 
@Preston_Bannister The SCARA advantage is just compactness. However, you don’t need or want a high speed machine for free printing. You can increase output by using a large nozzle which is appropriate for large functional parts.
@Nicholas_Seward I completely agree that SCARA has advantages for some uses (but not high speed printing). I also agree there is advantage in using larger nozzles. As my main interest is in faster 3D printing of larger functional parts, my experiments are along both dimensions.
Of late, experiments using larger nozzles are exposing the thermal limits of hotends, and hitting some hard limits. With a better map of the thermal performance of varying hotend designs, can better map when performance is due to limits on physical motion (which are in part mechanically limited, and in part limited by the software).
Take this as an exercise in eliminating unknowns.
I am sure you are aware of the high speed scara robots that you can see on youtube. They are fast. I believe that not speed is the limiting factor of scara but the materials and parts we use to make it into a 3d printer, and, most importantly, the speed with which one can spit out molten plastic is the truly limiting factor, namely the hotends.
@Florian_Ford yep. However, I think for a given sum of money one can typically make a cartesian bot with better speed and accuracy. The reason SCARAs are even relevant is that they aren’t bound to the volume of their frame and you can have more than one work in the same volume.
That said…I keep my interest in SCARA printers aimed at minimizing cost, maximizing portability, and enabling cooperability. So far with humble materials, the hot end isn’t a limiting factor.