And what exactly differentiates the commercial version from the open-source one? Is it that they took the lessons from their support & service business, and formalised them into a product?
I think the lesson here is that if open source software has nontrivial setup, there's money to be made by simplifying the packaging and distribution. Perversely it suggests you should make your open source software harder to integrate, so that people will pay you to set it up for them.
I think this lesson is there throughout the market - there are many companies now selling "X as a service" where X is something you can get and use for free, but there may be setup and maintenance costs.
Doesn't even have to be your software!
Worked for Redhat and OpenBSD...
The Guacamole case is really different since their customers will be by definition the IT departments. So, even if OS packaging was already good enough, you will still have a big chance of selling it to a big corporation.
But given enough time, isn't someone always going to come along and provide open source packaging for the project?
If so that would mean you're choosing to trade ease-of-install and community goodwill against revenue -- and revenue with a limited time horizon, at that.
I can see how that could appeal in situations where decision-making incentives are not aligned with long-term value of the overall ecosystem.
Selling services and expertise seems more sustainable, since offering those doesn't imply that you should add complexity or difficulty to your project. By providing expertise you can continue maximizing community goodwill and contributions, while reducing your own maintenance costs.
Providing services also enables the creation of an ecosystem around your project; that may include competition for your services, but it also implies a larger customer base and project adoption -- and more job creation and skill development.