Using a custom `glance` method in huxtable::huxreg()

I am looking for a way to report the number of groups alongside the number of observations using huxtable::huxreg to create a table of results of a multilevel model predicted with lmer(). I can write a custom glance method that overwrites the glance.merMod method from the broom.mixed package that works when called from R but does not work when calling huxreg().

I believe that the issue is that the huxreg function imports both broom and broom.mixed as required namespaces inside the function.

What would be the best way to overcome this issue to use the custom glance method (or an alternative other than simply adding rows to existing huxtables)?

Here's a MWE

library(lme4)
library(tibble)
library(huxtable)
library(broom.mixed)

## Simulate multilevel data
sigma <- 0.5
tau <- 0.1
x <- rnorm(100)
w <- rep(rnorm(10), each=10); i <- factor(rep(1:10, each=10))
y <- x + w + rep(rnorm(10, 0, tau), each = 10) + rnorm(100, 0, sigma)
d <- tibble(y, x, w)
m <- lmer(y ~ x + w + (1|i), data=d)

## Custom glance method
glance.merMod <- function(x, ret=tibble::tibble_row()) {
    ret$nobs <- nobs(x)
    ret$ngrps <- summary(x)$ngrps
    return(ret)
}

glance(m) ## Works, returns nobs and ngrps
huxreg(m, statistics = c("nobs", "ngrps")) ## Doesn't work, ngrps missing

You can fix this using tidy_override():

m2 <- tidy_override(m, 
        glance = list(
          ngrps = summary(m)$ngrps
        ), 
        extend = TRUE
      )
huxreg(m2, statistics = c("nobs", "ngrps"))

       ─────────────────────────────────────────────────
                                          (1)           
                               ─────────────────────────
         (Intercept)                            0.027   
                                               (0.064)  
         x                                      0.916   
                                               (0.063)  
         w                                      0.982   
                                               (0.075)  
         sd__(Intercept)                        0.096   
                                             (NA)       
         sd__Observation                        0.561   
                                             (NA)       
                               ─────────────────────────
         nobs                                 100       
         ngrps                                 10.000   
       ─────────────────────────────────────────────────
         *** p < 0.001; ** p < 0.01; * p < 0.05.     

You'll probably also want to call set_number_format on the ngrps cell.