Hi OnlyMe,
I don't know the details of how actual UTC standard clocks are synchronized with each other in practice, so this post is purely theory-based.
There is a theoretical problem with synchronizing clocks in a rotating reference frame.
Imagine a set of clocks spread around the equator. Starting from a master clock, each clock is synchronized with its neighbor using Einstein synchronization or equivalent. When we get back to the last clock, it will be out of synch with the master clock (its neighbor) by about 200 nanoseconds.
So, if we are to have ECEF clocks meaningfully synchronized to high precision without a discontinuity at some longitude, then whenever two ECEF clocks at different longitudes are synchronized, the synchronization has to be artificially adjusted so that they are synchronized in the ECI frame.
Chinglu's argument is essentially that:
1 - We should be able to use two close-together ECEF clocks to make measurements in their momentary comoving inertial reference frame
2 - Doing so would result in different speeds for East to West light and West to East light (by about 900m/s)
If these two points were correct, then this would obviously mean a violation of SR.
The second point is correct, but the first point is not. The side-by-side ECEF clocks are not synchronized in the comoving inertial reference frame, but in the ECI.
No, kidding? You mean if he knew how to do the clock synchronization he wouldn't make stupid assertions about relativity theory?