Should name= in compute.tbl_sql() use I() like in copy_to()?

[MichaelChirico]

dbplyr/R/verb-compute.R

Line 27 in 5726930

#' @param name Table name in remote database.

I noticed the new (#1456) recommendation to use I("schema.table") to indicate DBI::Id("schema", "table").

Does that also apply to other verbs such as compute()?

It seems so based on some new test failure I encountered when trying to update {dbplyr} that affected both copy_to() and compute() usages.

[MichaelChirico]

I see #1516 did explicitly just this in the NEWS.

Is it better to just add this note to ?compute, or should we aim to create a shared documentation of function arguments where I() is preferred & refer to that instead?