thin air… based on my prior quick test with cardboard and shrinking a 2x4x8" long to a 2x4x4" long and getting decent results. Looking mainly to shore-up the mechanics of something that seemed to electrically / RF work ok.
After lining the horn cleanly with Aluminum duct tape, I seem to be able to reliably get 3-4dB improvement. The +5 stated earlier must’ve been a fluke or over-zealousness in observations or less-than-optimal baseline measurement on the bare LNB.
I’ve done some EMI/RFI testing before, and note that the GHz frequency antennas tend to have a flared/parabolic horn. With 3D printing a lot more is possible than linear shapes. If there is a good online resource anyone knows about (eg online calculator) I could try some other shapes or angles, etc.
EDIT: Using this calculator: http://hornantennacalculator.blogspot.com/p/calculator.html
With the same 4" opening, an 8" length gives +20dBi, and a 4" length gives +16dBi. I would call that the same ballpark. Losing 4dB to make it half as big seems like a fair trade