It's known as industrial or scientific archaeology [1] and is a fascinating area of study. I found a good book on it, called 'Industrial Archaeology: Future Directions' [2] which most university libraries should have.
Interesting story and cool photos. Couldn't get to the end due to website design issues. Were these for horseback delivery or for airplanes?
Terrible, terrible UX in the form of a jittery delayed popover that fills the entire phone screen. If you are going to do this (and I can't stand the fact that it actually works on many people), make it so easy to dismiss and fast to load that I close it instead of instictively hitting back.
Air mail. These were the days before good roads crossed the country, and where even an automotive trip was an Adventure which would take weeks. Horseback, you're going slow enough that you'd be able to see much smaller signs, and would be on a pretty well-trod path.
Shortly thereafter, though, a few things happened that made the arrows pretty obsolete. Mapping of course, was a big one; you could now navigate a lot more reliably through unknown territory. More importantly, a network of radio beacons was set up. Charts had lists of radio beacons, with their frequencies. Pilots could tune in to hear them repeatedly chirp their identification in morse code, and use radio direction finders to set their heading accordingly.
There's one other feature that was developed in that time period which also made pilots' navigation job a lot easier. The federal highway system meant that there were good, very visible, roads, serving as routing beacons in their own way. Pilots would, and still do for a lot of general aviation planes, route close to highways, because they're also a very obvious landmark that carves a path through the country.
For the air mail planes that used to fly through the night. Most of those arrows were painted in luminescent paint, and had a light tower nearby shining down on it so that the arrow glowed and was highly visible to the poor cold, tired mail plane pilot above.
They did have compasses, but those do not give you your position. So if you've been flying for four hours at night, with a thirty mph wind, you could be 120 miles away from where you planned on being. You need some way of getting an absolute position, which these provided.
Windspeed is not unlike ocean currents, but since the speeds involved are a lot higher than in a boat errors in estimation accumulate way more quickly.
Not without aviation charts, which the article does mention were in short supply back in the early days of aviation. That, and I'm willing to bet that the licensing for pilots less than twenty years after the Wrights took off at Kitty Hawk was less than rigorous. So assuming that a licensed pilot, if such a thing even existed, was well-versed in navigation via charts and a compass was likely wrong.
I also find these kind of things cool. Forgotten structures that once served an important roll but now are mostly forgotten.