`prior_weight` parameter applied in VPCs?

[roninsightrx]

A user reported that runs where prior_weight was applied (e.g. 0.6, 0.3, 0.1, etc) resulted in exactly similar VPCs, which was surprising to the user. In a way, this is correct though: prior_weight is something that refers only to Bayesian estimation, and that's not relevant in the VPC. However, one could argue that prior_weight is just a modification (inflation) of omega, so if read that way it should be reflected in the VPC.

Perhaps the easiest way is to just keep the functionality as-is (doesn't affect the VPC), and issue a warning whenever prior_weight is not 1 and VPC is requested. E.g. an info message like "The prior_weight parameter is not equal to 1. This change in prior weight only affects Bayesian computations, but the factor is not applied in the VPC".