There's a bit of software I use in my daily life (completely unrelated to all of this) that's called World Machine, it's a terrain generation software. I bring this up because it is (as far as I'm aware) also only developed by one person, so I feel like it's a good point of comparison. For World Machine, there are a few different tier options for purchasing the software. There's the "Community" tier which is free, the "Indie" tier which is kind of a midpoint for the prosumer, and then the "Professional" tier, which is significantly more expensive. The important thing to note is that there is no real difference in core functionality between the three versions, the main limitation on the tiers lower than Pro is output resolution. Free version has a low maximum output resolution, mid-tier has what I would call a "high enough" maximum output resolution, and the pro version does not have a maximum output resolution. I feel like this could be a good framework to go off of. I'm not sure limiting the possible output resolutions for a theme would be possible or advisable, but perhaps it could be something as simple as limiting the number of views available in a theme for different price levels. To me, that strikes a really nice balance between incentivizing people to purchase the software to build more complete themes while also not being cost prohibitive to people who maybe aren't fully invested in the process (yet).