The light clock model assumes that distance is an invariant quantity.
When the quantity of space compressed within a given matrix system varies - so does the distance.
This is seen quite clearly in the phenomena of gravitational lensing.
No it doesn't. The distance the longer diagonal light path takes is measured in the stationary outsider's frame of reference, which of course is fixed with respect to him. The shorter straight up and down path is measured by the observer moving along with the light clock. It all stems from the assumption that the speed of light is the same for all observers, regardless of their state of motion.