How to obtain the posterior probability of 2D distribution?

Copulas are just schemes to draw correlated draws from arbitrary distributions. There’s no “better”, just different. It’s like asking if a Beta prior is more performant than a Normal prior.