The term ’length’ in GR is pretty complicated and there is no such thing as a canonical spacial distance between two events. There is no canonical splitting of

The term ’length’ in GR is pretty complicated and there is no such thing as a canonical spacial distance between two events. There is no canonical splitting of space-time into space and time, unless perhaps for very symmetric space-times. E.g. we use the free falling galaxies in an isotropic universe as a global clock and call the orthogonal complement ’space’, aka ’comoving coordinates’. Using the term ’speed’ in the sense of ’spatial distance per time’ implies some non-trivial conventions. So the most accurate way to describe light rays would be to say they are “light like curves” (the ones with zero velocity wrt the Lorentian metric) which are exactly the geometric entities that describe the causal past and future of events. So IMHO causality is the real physical property of space-time and the ’speed’ of light is just a convenient way to visualize it. ℏ is probably best described as fixing the units for angular momentum. The energy spectrum of QED is continuous. And using the properties of free particles feels a bit fishy as they are just an approximation.