I may get this half right as I'm not super hip to this stuff, but I have read that from a design and engineering standpoint humanoid robots are the least efficient possible whereas purpose-build robots (which typically look nothing like humanoid) are the obvious choice in most cases. Having said that, a lot of companies are chasing the humanoid form because they'll be able to immediately adapt to existing infrastructure (to replace human workers). They'll be able to occupy existing work spaces, use existing tools, etc etc. Which from a profitability/capitalism perspective is the goal.
The better these things get at doing general labor and even skilled tasks like harvesting crops, building houses, garbage collection, etc etc... the less jobs for the working class. Libertarians genuinely believe that AI and robotics will create a utopia but the self-same libertarian billionaire assholes will never stoop to paying taxes, or contributing to social safety net programs, etc etc. So for them robotics taking over as many labor markets as possible is a dream come true. They don't give a flying fuck what happens to the rest of us. If we don't have work, we don't have money to participate in capitalism. They're already outlawing homelessness and building concentration camps for the poor.
It's a good time to re-read The Grapes Of Wrath in my opinion.
I'm not saying I believe they can do it, or that even they believe they can do it. I'm saying that's the profit motive behind why they're trying to develop humanoid robots. If any of them can be "the first" to making ones that can use tools and spaces designed for us, they'll immediately get into "disrupting" labor markets etc.