Looking at the two trees from bird's view, it is so tantalizing to merge them. So many pieces seem to work together.
Yet, as Private User posted above, the crux is to reconcile the two "Mary Dinwiddie", the one said to be born before 1695 and the other said to be born 1707.
The before-1695 is highly reliable. There is a baptism record. But after that, things gets fuzzy.
To maybe help unlock this situation, (or maybe just for fun), let's brainstorm. Using the great material that NF and Martin posted, we can speculate:
> We have Mary-1695. She dies of young age before 1707 from mysterious circumstances.
> Mary-1707 is born. She is illegitimate. She may be an illegitimate daughter of her father, or her mother. I am tempted to say her mother because the father died in 1710 therefore the mother may have had a stronger incentive to protect her daughter. Also, there is a 3-year gap between Anne (1706) and Elizabeth (1709), thus Mary could have been born in 1707 from the same mother.
> To hide their shame, they gave her the identity of her older half-sister Mary who died earlier. Notice that we found no baptism record of Mary-1707.
> Now, this is why things gets so fuzzy from there. We have two people with the same identity. From then on, Mary-1695 and Mary-1707 are constantly confused. Historians and genealogists keep blending the facts about the two Marys.
> However, her half-brother Robert, Governor of Virginia, isn't fooled by the machination. He makes no mention of her in his will dated 1769. It may be that Mary is already dead. Or also possibly, Mary is still alive but the Governor wants nothing to do with his illegitimate half-sister.
> Mary-1707 dies in 1772. They engrave on stone that she was 65 year old. And why not? Mary-1707 was already disinherited and in her later years of life, why keep pretending she was 12 years older than she really was. Why keep claiming you are 77 when you actually are 65?
This is all speculation of course. But if there is some truth to this, it would explain why we have two "Mary" profiles that are so tantalizing to merge. And why the data about the Marys appears blended.
In conclusion, if we can build a story like this one with the evidence at hand, I don't think we should merge these profiles until we have much stronger evidence.