...didn't bother anyone until railroads entered the equation: astrology did take the difference into account, armies didn't move quickly enough to notice it: just keep adjusting to local time every day or so (even on horseback, any tactic is far more dependent on physical limits, so you're coordinating on the order of minutes, not milliseconds).

For railroads, even at 40 kph, an east-west train would be subject to time shift (not in the Lorentz sense, it would just keep switching "timezones" too quickly). Given that the primary security element was "this train is supposed to pass this set of switches between this and that time and wait there, else it risks colliding with that other train running opposite," both accurate timekeeping and geographically wider timezones were needed (as in "the railroad will use its own time, Prague's lunchtimes notwithstanding").

> That is true, but it did not bother anyone

Time zones were implemented in the United States in 1883. This was well after the use of telegraphs had become widespread, and after the invention of the telephone. While the telephone would not by able to make coast-to-coast calls for another few decades, the idea of communicating with people in other cities was already becoming commonplace.

My point was that the "messy" system the Times article describes would already have seemed reasonable to everyone, who already knew that people in other cities would have different working hours. What annoyed people was the "flattening" of the time zones, making it such that all the cities had to now synchronize their clocks to railway time, no matter what their local noon was. But this is the opposite of the point that the "one global time" advocate thinks he's making.