With powered reactors you're going for breaking the bubbles up instead of slowly letting them dissolve into the water column passively. This allows you to get high, consistent
ppm, especially since youre running pressurized. Venturi loop design has an open bottom whereas the bioball design youre looking at forces bubbles to break on the bioballs; through either design you won't see many if any bubbles escape, just
CO2 enriched water (though with venturi loop you won't want to inject
CO2 at a rate where it constantly goes past the burb tube).
Assuming you can replace the line should you want to sell the canister later, you can use the entire return line as part of reactor by injecting right after the canister: I do this with my sump return and have the same numbers I did with a traditional powered reactor, so I believe either method dissolves all the
CO2 my yeast based set-up produces. The effectively very long chamber forces bubbles to break up as its pushed along at a fast rate, but I only use
DIY in relatively small volume, and a bioball chamber may be necessary with pressurized at your volume.
Zezmo's idea of feeding the gravel vac with the canister outtake makes sense to me too, but I dont think you should use use a line reducer to fit the gravel vac's end cap and I think you may need a couple bioballs in there to break up the gas to get complete diffusion with your setup.
I understand this may be a PITA, but once you get complete diffusion or close to it you're done, and your caves make me sure you can do this easily
It's like 15 minutes tops for assembly, seriously.
HTH