Control of where Linux development is going is indeed economically worth a lot. Copies of the Linux kernel are near enough to worthless for the distinction not to matter.