And those costs won't disappear with GBAS. Not a single GLS system in the world providing ILS-like minima, and plenty of obstacles to overcome before it gets there. Won't be suprised if it goes the way of MLS. But we can be the expensive guinea pigs and provide a different solution to the rest of the world so Airservices can charge for the ground equipment.
We could have had WAAS providing coverage for the entire country at a fraction of the cost to provide the same with GBAS but Airservices kept pointing to the first version of WAAS in the states "look how much it went overbudget" well duh, they spent a lot of money figuring out how to make it work. But guess what? Now they've figured it out, it works and easy to reproduce! In 2008 there were over 1333 WAAS approaches in the US, with a plan to add another 500 a year. Try doing that with GBAS on the same budget. The more recent systems have been much closer to on time and on budget with the exception of one failed satellite launch.
But no, we'll install GBAS at the capital cities so we can cover "97% of RPT passengers", which could have (and probably will be) covered at zero cost and zero return to Airservices with Baro-VNAV instead, while the 3% that would have been covered with WAAS and won't ever get Baro-VNAV capability miss out completely (not many Baro-VNAV capable 404s around the place!)
A real shame, if we'd tacked WAAS on to the NBN satellites, it would have cost less than one GBAS install.