1. I agree with the "tastes like chicken" argument, especially for enterprise software. Most enterprise software is basically CRUD apps with a data analytics component where most of the complexity is in integrating with existing 3rd party systems and processes. It really is mostly plumbing.
2. I think this article does a fairly good job in raising the red flag that most of the profits are on paper. I say this because a funny thing about enterprise subscription software is that companies can spend a lot of money on it, but the purchasers are often not the people actually using the software. This can create a situation where things look OK for a while as the revenue rolls in, until one day when the folks with purchasing decision power actually talk to the end users, and the end users say how much they hate the product and all they are doing is working around it with an inefficient process (yes, I've been involved with enterprise software before). Then the software subscription is cancelled, and the subscription revenue can have a huge decline over a very short period.