Georgina (and Pentcho),
You have made some interesting observations which when pursued to their logical conclusion we could possibly arrive at same destination.
You asked the question "a second where?" and pointed out the various effects differences in gravity on Earth surface and at an altitude would have on the frequency count of Caesium 133 and thus the definition of a second.
Now, I ask further, "a second when?"
In early January, the effects of the Sun's gravitational field on our tiny blue dot (the Earth) is highest when the Earth is closest to the Sun,i.e. lower down in the Sun's gravitational Tower (perihelion), than in July when the effect of the Sun's gravity at the Earth's surface is lowest when we are at our highest altitude in the Sun's gravitational Tower (aphelion). That means even at Earth surface, the definition of a second will be inconsistent going by the gravitational effects on light frequency. This illustrates the illogicality of using a phenomenon to define rather than to give a reasonably accurate measure of time. Or can time be dependent on the time of measurement itself? Back to Newton!
Again, it is well demonstrated that a pendulum swings faster the lower it is in the gravitational field and swings slower at an altitude. This is directly opposite the frequency behavior of electromagnetic vibrations resulting in inconsistency of measuring time. Therefore, while the frequency of Caesium 133 is useful in keeping time and measuring a second to a high degree of accuracy, it is not the Time itself. Back to Newton !!
The conspiracy to use Light velocity and frequency to define Time is not unrelated to the quest to foist the constancy attributes of Time on light velocity (c), while exchanging the variable attributes of velocity under different conditions on Time. I attach a short paper of mine at the end of my post.
On the 'Faster than light' topic, you said, "Rate of production of the em waves is not altered by movement of the source or observer but the rate of receipt does change, giving the Doppler effect change in frequency". This means your ideas are opposed to Lorentz invariance, because in Special relativity, you cannot vary the time of receipt of a signal once it is emitted and already in flight towards you, but you can vary it before it is emitted by changing your distance from source. However, in Galilean-Newtonian relativity this is a possibility under some scenarios. Back to Newton!!!
Then for Pentcho's sake, when a particle is emitted from air and enters water for example, it slows down. On emergence back into air, it continues at that same slowed down speed or even less. Not so, for a wave that travels at a speed peculiar to the medium of propagation. Thus when light enters water from air its speed slows, but on emergence it acquires its characteristic speed in air. After interference of light was demonstrated emission theory lost relevance till it was re-introduced without full justification into physics, when other explanations for phenomena were plausible.
Regards,
AkinboAttachment #1: 3_CPEM_2014_Summary_Paper.pdf