As established by others, this is the unaccepted answer (twice!) to Essay in a sci-fi anthology; A missile from (circa) 1968 is found in 1940 - When do scientists think it's from?, "No Copying Allowed", an editorial (not a short story, really) by John W. Campbell, Jr., originally found in the November 1948 Astounding Science Fiction and available at the Internet Archive.
It's also set in the 1920s, and written in the 1940s (and then updated in the 1960s).
Let's first consider this situation: Time: About 1920. Place: An American Army Air Base. Action: High overhead a small airplane tears across the sky with a high, thin whistle. Ground observers, after tracking it for a minute or so—during which time it has passed out of sight—report incredulously that it was doing between nine hundred fifty and one thousand miles per hour. It circles back, slows abruptly as the whistle dies out, and makes a hot, deadstick landing. Investigators reach the cornfield where it landed, and find it ninety percent intact—and one hundred percent impossible. Swept-back wings, no tail, automatic control equipment of incredibly advanced design, are all understandable in so far as function intended goes. But the metal alloys used make no sense to the metallurgists when they go to work on them. The "engine," moreover, is simply, starkly insane. The only indication of anything that might remotely be considered an engine is a single, open tube—really open; open at both ends. But the empty fuel tank had tubes leading into some sort of small jets in that pipe. The athodyd being unheard of in 1920, the thing is senseless. Filling the fuel tanks simply causes a hot fire that must be extinguished quickly to prevent burning out the tube. The fact that this is a guided missile intended for launching from a four-hundred-mile-an-hour bomber makes the situation a little difficult for the 1920 technologists; the athodyd won't start functioning below two hundred fifty m.p.h., and nothing on Earth could reach that speed in 1920.
Discussing the updated version:
Since the drone originates in the late 60's, it incorporates ICs (although puny by today's standards), and Campbell comments on both the difficulty of detecting the impurity levels responsible for the chips' functioning, and the difficulty of monitoring such operation, since the oscilloscopes of the day were simply too slow to pick up logic signals.
I am satisfied that this is the story I remember being referenced.
