In contrast, the railway networks had been in place for almost a century by then. They were mostly privately-funded and are part of the capitalistic lore of risk-taking and revenue-gaining for businesses, as well as the historic lore of opening up the US west in city and country alike. Peter Drucker however said in a Claremont College lecture, that "the US rail carriers failed because they saw their business as rolling stock, whereas it was moving goods". Wittingly or not, trucking businesses seem to have gotten that distinction in Europe too.
Edward de Bono said in his book "Sur/petition" that the true winners will integrate all facets of a business, each step augmenting the previous and subsequent ones - the cited example was an auto manufacturer who also ran service stations and parking lots- the idea is that cars are not vehicles but a means to get from A to B, fulfilled by the combined services cited as an example. A recent example may be a venture into electric cars in Israel and Denmark by Renault and Nissan, which also secures sourcing of electrical outlets, and benefits from the short driving distances in those countries (the range of batteries has always been a shortcoming).
So it goes in the geospatial business too: If I can help users with their day-to-day business - to find petroleum in their data in my industry - then how I provide that service becomes less important. Not only does it free me up to provide whatever technology works best at any point in time - and we all know how quickly that changes - while users don't worry about that very technology. But it also helps me keep an eye on the ball, that is to provide users with tools that augment their workflows, rather than replace them. And if that means spanning technologies, then I'm better suited to address that too. And as above, the internet provides that externally developed and funded infastructure that lowers my barrier to entry, which in turn benefits the end users.