Using groups as table header in kableExtra - r

Question

I recently discovered kableExtra. Making tables in R instead of manually entering values in a Word-table is much faster and less prone to error.
I want to make tables that have a (or several) grouping variable(s) as a header.
Basically, instead of this
iris %>%
group_by(Species) %>%
summarise(mean = mean(Sepal.Length), sd = sd(Sepal.Length)) %>%
kbl(digits = 1,format = "pipe")
|Species | mean| sd|
|:----------|----:|---:|
|setosa | 5.0| 0.4|
|versicolor | 5.9| 0.5|
|virginica | 6.6| 0.6|
Instead I want to accomplish this, or a variation of this.
| | Setosa | Versicolor | Virginica | |
|------|--------|------------|-----------|---|
| mean | 5.0 | 5.9 | 6.6 | |
| sd | 0.4 | 0.5 | 0.5 | |
| | | | | |
For multiple headers, I was thinking something in the lines of this
iris %>%
mutate(long = ifelse(Sepal.Length > 5,TRUE,FALSE)) %>%
group_by(Species,long) %>%
summarise(mean = mean(Sepal.Length), sd = sd(Sepal.Length)) %>%
kbl(digits = 1)
|Species |long | mean| sd|
|:----------|:-----|----:|---:|
|setosa |FALSE | 4.8| 0.2|
|setosa |TRUE | 5.3| 0.2|
|versicolor |FALSE | 5.0| 0.1|
|versicolor |TRUE | 6.0| 0.5|
|virginica |FALSE | 4.9| NA|
|virginica |TRUE | 6.6| 0.6|
But instead producing
| | Setosa | Setosa | Versicolor | Versicolor | Virginica | Virginica | |
|------|--------|--------|------------|------------|-----------|-----------|---|
| long | TRUE | FALSE | TRUE | FALSE | TRUE | FALSE | |
| mean | 5.3 | 4.8 | 6.0 | 5.0 | 6.6 | 4.9 | |
| sd | 0.2 | 0.2 | 0.5 | 0.1 | 0.6 | NA | |
| | | | | | | | |
Bonus points for not repeating the table header, but having a merged cell.
Can anyone point me to any examples or relevant documentation?
I am using R 4.0.2, and as such only have access to kableExtra, not kable.