Done in spglm/#7, should allow aicc to be calculated for GLMResults object.

Thanks, just pushed a new commit reflecting that.

Looks good. A few potential tweaks:

1. Will the method work for Poisson/Binomial? They have slightly different model fit diagnostics
2. For MGWR, Should we add the corrected alpha value and corrected critical t-val for each variable next to ENP_J? This could be good to report.

For 1, yes, it works for Poisson/Binomial, see here at the bottom, I attached a summary output for the tokyo data.
For 2. Added the 95% covariate-specific t and alpha for mgwr in the new commit, also see this in the mgwr output.

Also, here is the output from gwr4 for the tokyo one.