There's a really simple two-part recipe for failing to create useful ubiquitous services at universities. It goes like this:

1. come up with an important service that should be ubiquitous.
2. fund it by charging its users.

The moment you start charging people for things at a university, you enter into a competition for their money. People start asking 'why should we spend money on this?' and 'why should we pay you for this?' and 'is this the best use for this money?', and you can't win that competition all the time.

(Some unit is always strapped for cash and will opt out. Some unit always feels that they can do it better than you, or just that their money is better spent on something that fits their needs more precisely.)

In short: because you've given people a choice about it, you invite them to not buy into the service. So much for ubiquity.

(Also, simply doing the sales job drains resources; you wind up spending at least some of your time selling people on the service instead of providing it.)

With central budgeting, you can theoretically give people money with one hand and then mandate that they spent it with the other. This probably isn't going to be popular (partly for reasons outlined here), plus it doesn't necessarily work all that well once you hit people with grant funding.

If something is important enough to be ubiquitous, the only way to go is to give it away for free. Don't try to fund it on the sly by charging people on a cost recovery basis; persuade the central budget authority that it's important enough to fund centrally.

This even applies to purely nominal charges; sooner or later someone will decline to pay, and you lose the benefits.

(Disclaimer: I am pleased to say that the University of Toronto gets this, and things like Windows anti-virus programs are free for anyone on campus (and funded through central budget funds).)