lists:reverse/1 vs. lists:reverse/2 performance

[TD5]

Describe the bug

lists:reverse(List,[]) is ~50% faster than lists:reverse(List) for big lists, and about the same for tiny lists, according to my benchmarks.

Unlike lists:reverse/1, lists:reverse/2 is implemented in a BIF, and is documented as follows:

Returns a list with the elements in List1 in reverse order, with tail Tail
appended.

Example:

$ lists:reverse([1, 2, 3, 4], [a, b, c]).
[4,3,2,1,a,b,c]

Expected behaviour
Comparable performance between the two.

[TD5]

Hmm, after rerunning my benchmarks with more samples, I can't reproduce this, so I guess it was just noise.