Is it worth switching from dplyr code to data.table in r shiny that have various reactive data?

I am sort of new in R recently came across data.table package that is probably ~ 10 times faster than dplyr in various operations.

I have a shiny app based on Covid data that is getting heavier each day I am already not so impressed with the loading time expecting it to get slow by each passing day.

In shiny app I have provided several input options to user hence the reactive elements and filter select operations in it.

Advice Needed on:

1. So is it worth taking the pain of changing the code from dplyr to data.table to improve the performance ?

2. Will it improve performance in shiny as well or there are any conditions ?

3. Is data.table compatible with shiny or reactive elements etc. ?

And this could be the main reason for me to start learning data.table.

data.table Benchmarks I came across: https://h2oai.github.io/db-benchmark/