We experimented with a few different form factors, including integrating into existing game engines, before taking this path. Our rationale was that we wanted to go beyond the existing engine integrations (i.e. Unity already has an AI assistant) and reimagine the entire game creation experience with AI and multiplayer in mind.
Our platform's entirely on the web, so you can link your current game/game creation session to someone else and have them join you. This works on most modern browser, making it easy to build and share games.
That being said, you're not wrong that the current scope of games is limited. That's OK with us for now - we'd like to focus on getting the core experience as good as possible with the current scope and then expand outwards, instead of trying to build something that attempts to solve everything and works for nothing.