`cross::run(pkgs =)` could take a data frame for `pkgs`

[DavisVaughan]

df <- expand.grid(
  vctrs = c("vctrs", "r-lib/vctrs"),
  rlang = c("rlang", "r-lib/rlang"), 
  stringsAsFactors = FALSE
)
tibble::as_tibble(df)
#> # A tibble: 4 × 2
#>   vctrs       rlang      
#>   <chr>       <chr>      
#> 1 vctrs       rlang      
#> 2 r-lib/vctrs rlang      
#> 3 vctrs       r-lib/rlang
#> 4 r-lib/vctrs r-lib/rlang

Doesn't make sense for run_branches() because it is contained within a single package

I think the $pkgs column in the return value would just be a df-col containing this