How can you possibly say that speed has nothing to do with the lenght of a second???
Lets make this easier to understand. Sunsecond stands for a second in solartime. Atomicsecond stands for a second in atomic time.
Untill 1956 a sunsecond was the 1/86400 part of a sun-day. A sun-day is the time between the sun reaching its highest point in its orbit and the next time this happens. Take a look at Wiki to see that this is correct.
I think we can now safely say that the lenght of a sunsecond depended on the lenght of a sun-day untill 1956.
There is now data available that shows that not every sun-day is as long as any other sun-day. The lenght of a sun-day is not constant. There are indications that sun-days were shorter in the past.
If sun-days were shorter in the past and a sunsecond is a 1/86400th of a sun-day. Then a sunsecond was shorter in the past as well.
Any speed that exists has something to do with a distance and a time it took to cover this distance.
Time is always involved in speed. I don't see how that can be denied.
The measurement of time seems to have been more or less incorrect in the past. Therefore the measurement of speed has been more or less incorrect as well.
If I am mistaken here, please show me.