If ligth's path is not euclidean then we could have a Copernical Observation Error?

A popular belief is that space is non-euclidean.

An "exagerated" example of a non-euclidean Copernical Observation
Error:

If light's path is not straight(non-euclidean) and bent perhaps in a
way similar to the Earth's surface then I believe that far away
objects would be observed more distant from each other then they
really are? (Pretend your standing at the North Pole and all of
Earth's longitudinal lines are ligth's path to your eyes)

Therefore farther objects would be seen to expand from each other
less, and closer objects would be seen to expand faster (since the
rays of light are more condensed at closer proximity) ....which is the
current observation through the HUBBLE telescope (farther objects are
expanding less then closer objects).

-------------------------------------------

If we could see everything then the south pole (a distant star) would
be observed from all directions in the North Pole and this does not
concur with Hubble Observation.....but if we couldn't see "THAT" FAR
AWAY, then perhaps..??

.