From census records, I've worked out that mine was potentially born early April 1808.
The 1851 census was taken on 30 March 1851, and George said he was 43, which, if accurate, would mean he was born between 31 March 1807 and 30 March 1808.
The 1861 census was taken on 7 April 1861, and George said he was 53, which, if accurate, would mean he was born between 8 April 1807 and 7 April 1808.
The 1871 census was taken on 2 April 1871, and George said he was 62, which, if accurate, would mean he was was born between 3 April 1808 and 2 April 1809.
As there's no overlap between his putative dates of birth from the 1851 and 1871 censuses, at least one of them has to be wrong, and your best estimate is that he was born between 1807 and 1809.
This doesn't fit well with the baptisms of the children of George Henderson and Ann Mitchell.
There are 18 months and 20 days between the baptisms of George on 23 February 1807 and John on 12 September 1808, and only 14 months and 19 days between the baptism of John and the end of 1809.
So I think you are right to be wary of the Anc* trees (actually, it is
always right to be wary of
any online tree!).
One thing you can do, if you have access to Anc*, is to look at all the trees there and see what source they cite for George's birth/baptism. Dismiss every tree that cites another tree or a secondary source (e.g. another web site or a book written long after the event), because you know that their authors have just copied someone else, and then see what you are left with. If they don't cite a primary source they cannot be trusted.