Path-dependence and the lock-in of inefficiencies and errors suggest beguiling alternatives to market-based explanations. Trouble is that the evidence presented for lock-in is weak (Liebowitz and Margolis).
The simplest example must be U.S. settlement change. The Europeans made a mistake and originally landed on the wrong North American coast. That error has been relentlessly corrected over the years. The Census Bureau's mean and median centers of population have been steadily moving southwest for years; the sunbelt-frostbelt migrations are well known; empirical models of migration consistently show that climate matters.
Do people move to places with warm summers, warm winters, or moderate variance between the two? Looking at 1990-2000 city growth, hot summers explain best. These places may also be attractive because they were also cheap. Perhaps because they had been hot and undesirable before cheap air conditioning and affordable swimming pools.
Be that as it may, people are very busy correcting the error that the Pilgrims and many others made, starting in the early 17th century, landing on the wrong coast -- and very far away from the left coast.