2. Links between the children listed as those of GWK are in the majority no more than best guesses based on names, ages, locations and professions. Some are more certain than others. In decreasing order of confidence I would places these as:
George Whitfield - 100%, 1884 reference to his job application as postmaster to succeed his father.
Andrew Guyse - 95+%, named for GWK's father
Farrar - 80+%, names of his children
Lawrence - 70+%, witness at the wedding of Frances Elizabeth (AGK's daughter)
Thomas - uncertain, age and location only
Eliza - uncertain, age and location only
William - uncertain, age and location only
Sarah - uncertain, age and location only
Much of the information I've used came from Roger Fisk, great-grandson of Farrar Kinsman, which I have then added to and cross referenced to the sources that I've found.
3. The DNA test is completely uncertain until you've had it done. Sorry, but there is no way of predicting of whether it will come up with a direct match with one that's been done already or turn out a completely new result. Even if it is a new result then it is not wasted. The first person to have their's tested in our group was a Kinsman in the USA - no match with anybody until about five years later when I persuaded a cousin of mine (Kingsman from New Zealand) to have his done. They matched 32 markers out of 37 and proved a link that none of us believed before.
4. Very good point about you and your Dad. Quite pointless to check both of you and better to have his done rather than yours because his is one generation earlier than yours and therefor has one generation less opportunity for changes.
If you do decide to to it, then I'd suggest that you join our Kinsman group at ftDNA, not that it's hugely cheaper, but it will save you $20. The link is:
Derrick