Interesting to note there that initially trying to use the short stick to directly retrieve the reward is considered a mistake; I wonder how many humans would give that a try, though.
My guess is most humans run two quick analyses: 1) what is the cost of seeing if the short tool can be made to work and 2) what is the cost of retrieving the more distant long tool using the short tool. Human/primate efforts would seem to begin with tool use before escalating to metatool use.
All seven of the New Caledonian crows in the experiment engaged in metatool use on their first attempt; only some of them successfully retrieved the second tool but none of them trialed simple tool use.