Fix data frame initialisation in bind_rows()

[lionel-]

Closes #5417.

[lionel-]

Performance impact:

new_list <- purrr::map(1:10000, function(x) list(A = 2 * x, B = 4 * x)
bench::mark(bind_rows(new_list))[1:8]

#> # Master
#>   expression               min   median `itr/sec` mem_alloc `gc/sec` n_itr  n_gc
#>   <bch:expr>          <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl> <int> <dbl>
#> 1 bind_rows(new_list)    181ms    194ms      5.25     821KB     14.0     3     8

#> # This PR
#>   expression               min   median `itr/sec` mem_alloc `gc/sec` n_itr  n_gc
#>   <bch:expr>          <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl> <int> <dbl>
#> 1 bind_rows(new_list)    552ms    552ms      1.81     821KB     9.05     1     5

[hadley]

And #5749

[lionel-]

@DavisVaughan Good point about name-repair. We don't need to repair names before calling vec_rbind() since vec_rbind() repairs them. Fixed in 827615c.

The perf example above was reduced from 370ms to 240ms. Thanks!