Some thoughts on the topic:
A foundation does not have to be useful for everyone. If it's useful for someone, then it is already useful. And just because Lua has being doing fine without a foundation, it also doesn't mean a foundation isn't useful or needed. A foundation can still add value. Good things can improve. And there are many needed things that are out of the scope of LabLua.
Below are just some of the things where I can see a foundation being very useful:
* Serve as a slot and point of contact for companies who wish to sponsor and support Lua related projects that are not necessarily Lua itself
* Develop things that are outside of LabLua's scope of implementation of the language, such as improving the ecosystem and developer experience. This could include things like maintaining critical modules, creating plugins for common text editors...
* Work with community leaders and maintainers of critical tools. E.g. See if the LuaRocks folks need help with anything.
* Advocate and sprinkle more Lua around the world with things like coordinating events, managing social media, writing blog posts, tutorials or other online resources...
This is not the first time a discussion around the possibility of a Lua Foundation came up. That I saw, there were discussions on this topic at least during the past two editions of the Lua dev room at FOSDEM. I've even brought this subject up with Roberto in the past, and others have done the same. Some folks are really keen on seeing something like this happen. I'd be one of them :)