In terms of the raw financial case for overbuilding now to avoid expansion in 15, 20, 30 years essentially depends on the discount rate used by government. That is, what kind of `return` are we looking to get on infrastructure spending. Many businesses use discount rates in the 12 - 15% range, which makes overbuilding for the future much worse than building to demand now, and upgrading in 15 years even if it might cost 5x more at that time.
It`s arguable that government can use much lower discounting, since it can borrow for so little. You still need to factor in something more to account for the risk the investment may not be needed for 50 years, or ever. I don`t know what rate, if any, is used by government in their CBAs. My guess is that something in the 7-8% range would be appropriate. At that rate, it still doesn`t make sense to overbuild now, even if it costs one tenth of what it would cost thirty years from present.