I'd never heard of Sanderson prisms, but the papers you reference suggest they may permit very low-cost DIC photomicrography--apparently a track you're pursuing elegantly.
Are the apparent shadows in your first onion image from the Sanderson prisms alone, or did you partially occlude the condenser?
I notice that eBay has cheap listings for polycarbonate (aka Lexan, Macrolon, Tuffak, etc.) from U.S. sources, with free, expedited shipping. (Such as here and here). The 1/4 inch thickness matches the 6.25mm that Sanderson used, and is close to the 6mm named by Schultz et al.I got a thicker (5mm) plate few days ago (Aliexpress, very long shipping)), but it is too big for this setup, have to find a nice way to cut it into the pieces.
I'm inclined to order a piece to try myself, though I'll likely mill an aluminum fixture for it, rather than 3D print something. Not that milling is better, but it's more familiar to me. Also, if I try the approach, it will likely be with reflected, rather than transmitted DIC--and reflected DIC is likely to be much less exciting.
From a bit of reading, polycarbonate appears to be easy to cut with a saw.
Please keep us updated with your further experiments.
--Chris S.