The earth is near the center of the universe, at the bottom of a deep
gravitational well. Relativistic effects result in billions of years
passing in the rest of the universe while only thousands pass near the
earth. This explains how multibillion-year-old stars and galaxies can
exist in a universe only a few thousand years old.
Gravitational time dilation, if it existed on such a large scale,
should be easily observable. On the contrary, we observe (from the
periods of Cepheid variable stars, from orbital rates of binary stars,
from supernova extinction rates, from light frequencies, etc.) that
such time dilation is minor. There is some time dilation corresponding
with Hubble's law (i.e., further objects have greater red shifts), but
this is due to the well-understood expansion of the universe, and it is
not nearly extreme enough to fit more than ten billion years into less
than 10,000.
Humphreys tried to use clocks in the earth's frame of reference. But
the cosmos is much older than the earth. Judging from the heavy
elements in the sun and the rest of the solar system, our sun is a
second-generation star at least. Billions of years must have passed
for the first stars to have formed, shone, and become novas, for the
gasses from those novas to have gathered into new star systems, and for
the earth to form and cool in one such system. The billions of years
before the earth are not accounted for in Humphreys's model.
Humphreys's theory assumes that the earth is in a huge gravity well.
The evidence contradicts this assumption. If the earth were in such a
gravity well, light from distant galaxies should be blue-shifted.
Instead, it is red-shifted.
See Conner and Page (1998) and Conner and Ross (1999) for several other
technical objections.
There is a great deal of other independent evidence that the earth is
very old.
If there were any substance to Humphreys's proposal, at least some
competent cosmologists would build on it and share in the Nobel Prize.
Instead, they dismiss it as worthless.