I'm more inclined to have casters in their own category instead of players vs. distributors. As leelin mentioned, casters add their own value by adding spit and polish to the games, which increases the reach on the games itself. Distribution platforms tend to not have any hands in the content but benefit more from getting a good "flow" of content through their network. As for monetization though, in every industry, there are outliers in each category that make a lot more money but in general, (successful) distribution platforms generate the most nominal revenues and frequently the people adding polish don't. The more skill/polish added though, the higher the premium it seems.

I'd like to echo seeing things through to a prototype. Unless theres some hurdle (technical, time, etc.) actually standing in your way, I fully support the idea of finishing the product in one form or another.

When you start questioning the idea, focus on the use case you're writing it for and cut out all the excess features that don't contribute to it. This will help you get closer to completion.

Remember that you're just one user and cannot possibly be representative of the population. Go test it to see what the market thinks!

As a product manager in a social gaming company, I definitely kept an eye on those metrics.

However, I'd like to caution that these metrics are relevant in all phases of the product lifecycle, but the focus on feature development follows the order you mentioned.

For example, keeping an eye on Rev/DAU and conversion rates at the beginning of the product tells you about user engagement, while paying attention to reach metrics (virality and new customer acquisition) in the mid to end will give you an idea of longevity or potency of your user base. Specifically, pay attention to social/user metrics as they will tell you how active a user is in interacting socially as well as inviting new users to use your product/game.