It can only be inductive effects, and I think maybe you're underestimating their strength Rad. Given that Et3N is ~1.5 pKa units more basic than NH3, each alkyl group contributes ~0.5 pKa units, which corresponds to a 5-fold shift in the equilibrium. So inductive effects are clearly able to overcome the influence of simply having a second nitrogen (which should only double the rate, to a first approximation).
I realise I am using a thermodynamic metric to try to explain a kinetic behavior, but I think the inductive HOMO-raising effect of hyperconjugation must be noticeably greater in quinuclidine vs DABCO, since the sigma bonds are much more electron-deficient in DABCO.
Either that, or it's some weird solvent effect, in which case god help us.