I think this is a case of confirmation bias. :-) There have been and continue to be many very good on-premise enterprise software products. And conversely, there are many not-so-good cloud products. I have worked on both at the enterprise level and have seen winners and disasters in both camps.

The factors you cite in your post definitely mean that the potential, and over time, the likelihood of SaaS products being “superior” (let’s assume that is easily measured) to on-premise product is real. But there are many factors that go into a good product that have little to do with the factors you cited.

These include effective customer and user research (meaning the research is done well AND proper insight understood), design (applying those insights into the DNA/architecture of the product), ongoing learning and iteration of customer/user needs outside of the product (the product is not an island unto itself) etc.

Great products come from understanding important problems and objectives and delivering great solutions to them, and then rinse/repeat. Data, APIs, etc. all can help, but they are not the reason that great products are created and delivered.