r2_zeroinflated() for zero-inflated negative-binomial count model?

[annacecile]

Hello,
I'm fitting a zero-inflated negative-binomial (zinb) model for count data, originally in glmmTMB.
As the performance package did not work for this model, I fitted my model again using the zeroinfl() function from the pscl package. Here the performance package works - but I have further questions on its calculation of r2:

1. Is the r2_zeroinflated() a sensible goodness-of-fit measure for zero-inflated negative-binomial models (as the example in the package description is a zero-inflated linear model)?

2. Is there a reference this r2 is based on (none included in the description)?

Thanks a lot!