The Washington DC Metro system is a busy rail system that has become an essential part of DC transport. Yet it is at capacity in terms of vehicles and the system is starting to break down. There is a bias in transport finance towards large capital expenditures and very little interest in maintenance. This isn't limited to the US (the London Underground is facing capacity constraints and dilapidated facilities) nor transit (the Interstate bridge collapse in Minneapolis was a maintenance/replacement issue).
With any technology the bulk of the economic benefits come after the technology is mature, not when it is new. Yet we finance expensive new systems without adequately thinking through how we will pay for them over their useful economic life. Relying on large federal payments is politically difficult and a poor use of federal dollars in many cases. Sales and use taxes to pay for infrastructure maintenance is regressive, and too often maintenance projects do not offer good enough photo ops for politicians (with the exception of an occasional pothole filling, though here in the governor California digs his own potholes to fill).
User fees seem much more fair and efficient, though there figuring out the right user fee is difficult. High transit fares will discourage riders. Road tolls are a good solution especially in areas where the potential revenue from tolls is substantial, though they are extremely difficult to pass politically. Higher gas taxes are okay, but as the price of gas rises people will shift to other modes or more efficient vehicles, lessening the financial benefits.
Somehow travelers need to pay for what they use through some type of marginal cost pricing. To say that new taxes, fees or tolls are bad because we've already paid for infrastructure and facilities through gas or other taxes misses the big picture and main problem we face. We are not paying for the infrastructure we have. We are only paying for part of it, and not even the part where we derive most of the economic benefit, namely the technology at maturity.