While not yet sufficiently playable to use to play music, this proof-of-concept folded contrabass recorder will produce a tone and gets within a couple of steps of the intended F2 fundamental pitch (88 Hz).

The bore diameter, length, vent sizes, and vent locations along the bore length were simulated with INRIA's OpenWIND python module. The finger-distance and print-volume constrained bore route was produced by my BoreOpt.m MATLAB script. This GitHub repository has the source code for these:
https://github.com/thecowgoesmoo/openwind

Actual mouth height, edge, and windway dimensions for the mouthpiece are not currently modeled by OpenWIND. Further, I have not had sufficient success designing recorder mouthpieces solely from methods like Ising's number (which gives an indication of which harmonics of a bore will be available for given windway+edge geometry). Forced to still use some iterative design, the mouthpiece included here is modular. The windway or ramp may be printed alone and then placed into the larger mouthpiece housing that mates to the bore via the double-conical union.

There are six holes, yielding whistle-like fingerings. They are intended to be reached with thumb, index/middle finger, and pinky of both hands. In this version I can manage the reach for the left hand, but the right hand is a bit larger than I can manage. A future version may remedy this with a different bore routing or with vent stems to reduce the required reach.

I haven't simulated the performance for larger vent holes, but for this bore size and pitch range, there may prove to be a significant benefit in bore admittance (inverse of impedance) for vents closer to the size of the bore. Such vents would be too large to cover with fingers, but simple printed keys like the parametric one I have published may be suitable for the task:
https://www.thingiverse.com/thing:6484695

I'll record a demo video and include a link to it here once I have the instrument playing well enough to demonstrate.