Adding a label to stat_mean in ggscatter

I'm looking at ggpubr::stat_mean:

Is there a way to add a label on top of the group mean point indicating the groping label cyl? Thanks!

[reproducible example:]

# Load data
library(ggpubr)
df <- mtcars
df$cyl <- as.factor(df$cyl)

# Scatter plot with ellipses and group mean points
ggscatter(df, x = "wt", y = "mpg",
          color = "cyl", shape = "cyl", ellipse = TRUE)+
  stat_mean(aes(color = cyl, shape = cyl), size = 4)

It's a bit unclear to me how you would do that with ggpubr. For some reason it doesn't seem to like to let you use the geom_text with stat_mean.

One work around is just to use ggplot directly with a bit of dplyr to calculate the means. You can do

library(dplyr)

means <- df %>% 
  group_by(cyl) %>% 
  summarize(across(c(wt, mpg), mean))

ggplot(df)  +
  aes(x=wt, y=mpg, color=cyl, shape=cyl) + 
  geom_point() + 
  stat_ellipse(aes(fill=cyl), alpha=.2, geom="polygon") + 
  geom_point(size=4, data=means) + 
  geom_label(aes(label=cyl), color="black", nudge_y =2.5, data=means) + 
  ggpubr::theme_pubr()

to get