allow bicop_dist() with tll

[tvatter]

What's the usecase for this?

[tnagler]

Had someone thinking that hbicop() doesn't work with tll because the hbicop(..., family = "tll", parameters = ..) API doesn't, which is the part that is most prominently documented. I couldn't see much of a reason to disallow it.

[tvatter]

Got it. Yeah the usecase makes sense. The only thing that worries me is that the check of whether it's a copula is pretty mild no? But someone constructing manually a tll is likely a sufficiently savvy user.

[tnagler]

Yeah I kept it intentionally mild to minimize false positives. But if someone does something like 4 x copula density or normal margins, it will detect it.