What you're saying seems like a potential false dichotomy; developing open source software can be as profitable - sometimes more profitable - than developing proprietary software.
Roughly speaking: focus on producing something of value, and then the monetization strategy should follow naturally if the project succeeds.
At the same time, I think it's important to focus on reducing total cost of infrastructure, because doing that increases the size of the audience who can participate and reach a sustainable value threshold.
(and one caveat to that: long-term cost is often minimized thanks to high-quality, durable systems that often have high associated price points. it may be necessary to have entry-level alternatives available too, although I don't think that's an excuse for providing poor-quality systems, or for entrenching high price points on legacy trusted systems)