Well the eye is never really completely still.  Even during fixation there are muscle tremors and other miniature eye movements moves the eyes slightly (Carpenter, 1977).  For convenience, at this moment, I will refer to all of the miniature eye movements as tremor.  One way to think about this situation is that  the brain may average the varying inputs at a higher level.  This feature would be easily accomplished by temporal summation.  The presentation of the model below illustrates what the this type of averaging might look like.  The averaging is a weighted average over the range of the eye tremor or jitter.  You can manipulate the range of this tremor below and see the output.  You manipulate the range of the tremor by indicating the radius of the tremor.  The current limit is 8.  You can enter larger numbers but it will not change the output from that of 8.  The receptive fields and the inputs can also be manipulated as before.

None of these manipulations seem to very helpful, at least as far as I have seen.  If you do find something suggestive let me know.  The general effect of the averaging is to smear the edges and weakens the effects of the lateral inhibition.  On retrospect there is no a priori reason for the higher levels to sum more slowly than the retina.

Another way to view the effects of eye tremor is to animate is and watch the variation over time.