Thanks for the suggestion. I have a couple of different modeling tools I use and the one shows B4 as 3 cf @ 44 for f3 of 44; Bessel as 2.5 cf @ 35 for f3 of 53. When I run WinISD I get a slight bump (only about 1db) and then a faster drop than you are getting, but it isn't bad. (Using Fs 51, Qts .47, Vas 2.4) A 3" duct would run 6-8" depending on the tuning and could easily fire out the bottom of the box.I think of this as overdamped in the same sense that increasing size in a sealed box drops system Q. I have used this approach before in bookshelf systems. For instance dropping Fb from 55 to 42, knowing I wouldn't be sacrificing that much power handling above Fb and that I would now be safe on low 'E'.
Dan