While $184M may sound like a lot, compare that to the $200M+ that was spent on the failed launch of Healthcare.gov - taxpayers got nothing out of that. That work was scrapped. That website ultimately cost closer to $840M.
What you don’t see in these numbers are the insane requirements needed to launch successfully. It’s not just building an AirBnb clone for parks, there is a huge amount of bureaucracy and stakeholder management in a national project like this.
$20M in annual recurring profit also doesn’t sound like that much in the grand scheme of things. Morally I wish it wasn’t necessary, but practically it’s impossible to deliver quality customer service without some form of financial performance incentive.
Would I love to live in a world where government employs programmers and DIYs this stuff much cheaper and more efficiently? Of course. USDS and 18F are bright spots that are trying. But they also don’t have the capacity to work on anything except high priority projects.