But we are not born from spherical balls of gas.
Correlation is not causation indeed, but sometimes it is.
But, I ask again, how would you prove this?
Freud (and you) are making a factual claim. To be a factual claim it must be falsifiable, that is there must be something I can measure which has the potential to show the claim to be false.
This is the problem with Freud, it is all very elegant, but there is no method of falsification - all he provides us with is story and myth.
If you could show me that, for example, people with absent fathers are more likely to believe in God, or that there is less belief in God in matriarchal societies, or that people angry with their fathers are more likely to be angry with God - then I might be slightly more sympathetic to your cause, though none of these would be even close to conclusive.
You can believe what you want, but personally I reject Freud's theories for the same reason I reject God - a complete lack of falsifiability.