It's not a question of speed, due to current computing limitations sequential search is a lost cause.
if you have 1000 petakey/s
and you search in a range 50% greater than your speed is not feasible -R + sequential.
because your software would not influence the first digits.

on the other hand, a real random has more opportunity with less computing power.
example

pk =1000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
your code                                                                                          100000000000000000000000000

random sequential just changes the values in color "red" more slowly, than with a true random.