Make usage of across() and data frames in filter() defunct

[DavisVaughan]

Closes #7758

Deprecated in 1.0.8, has been warning on every usage since Feb, 2022

I've added a special error message if we detect a data frame in filter(), as that is probably an across() call that should have been if_any() or if_all(). It may have false positives, but the message is crafted in a way that that is expected.

filter(mtcars, "x")
#> Error in `filter()`:
#> ℹ In argument: `"x"`.
#> Caused by error:
#> ! `..1` must be a logical vector, not the string "x".

filter(mtcars, across(mpg, is.na))
#> Error in `filter()`:
#> ℹ In argument: `across(mpg, is.na)`.
#> Caused by error:
#> ! `..1` must be a logical vector, not a <tbl_df/tbl/data.frame> object.
#> ℹ If you used `across()` to generate this data frame, please use `if_any()` or
#>   `if_all()` instead.