1. The project must have the product already working.
2. The project is supported by a team of developers and advisory team who have a clean and good track record.
3. The project has a detailed whitepaper.
4. Long-term plan.
I like your list above, but the issue I have with most of these project with working product is overvaluation of their tokens, if they have a ready product why not just move to the market straight but they will first come for ICOs to milk investors dry.
When you look at the lifecycle of a startup the minimum viable product (mvp) is rarely the end product. It is often used to show that the project is doable and that people are interested. With the initial funding the product goes through a major overhaul to meet functional and non functional requirements. Security and scaling issues must be dealt with and everything must be rigorously tested, user interfaces evolve, etc. Look at something like Instagram - it looks almost totally different from their initial 'working product'.
Succesful products also need marketing which requires a large amount of funding. Great products, even working and refined ones, won't sell if people don't know about them and relying on word of mouth by itself can take a long time to generate results.