KateWild:
If a man beats his child for 3 yrs at 6months intervals approx, is it probable he will beat his son in the next 6 months? I would say it is probable and allow the child to live with his mother as he wishes. It's not testable, and nor should it be, but the probability of beatings recurring is high.
That hypothetical example isn't remotely similar to your assertion about 'carbon and God'. It's based on a specific (but hopefully fictitious) sample of data, and it is testable. But probably better not to though.
Carbon is complex and beautiful, so I think of God. There is not need to test my theory, no one has proven my theory wrong, but I am open minded to the possiblity and it fuels my enquiry into science further to test my theory or prove myself wrong.
'Beautiful' is entirely subjective. 'Carbon is complex and beautiful' is meaningless as a data sample for testing probability of the existence of God. As already stated, it is not possible to test your 'theory'. But if there were a testable theory for the existence of God, given all the claims Christians make about how important it is to believe in God, there would certainly be a 'need' to test such a theory.