The thought process in my last posting was "corrupted" and should be ignored.
Nothing better than a test.
a: Gedcom created in RootsMagic 4 loaded into both FTM2011 and PAF5.
b: PAF 5 file loaded into FTM2011
1: Loading the PAF5 file generated 43 errors. These were all unrecognised dates, but in reality only 1 should have been a problem.
2: The size of the FTM file generated from PAF was in line with my expectations.
3: Looking at my own entry seemed to show no problems with relationships across the 5 generations displayed.
4: I then looked at the relationship to my Home Person which reads "Husband of 17th great grandaughter of Richard Wherrett". There is no direct relationship to the Home Person for my late wife. The actual relationship is very complicated and in my RootsMagic gedcom load into FTM reads as follows "Husband of 4th cousin 1x removed of husband of step 2nd great grand daughter of wife of 9th great grandson of Richard Wherrett"
5: If you follow up my wifes line you should arrive at a Richard Werrett who died in 1750 in Yate in Gloucestershire. His parents are currently unknown as I have a number of candidates, but to link together another branch which is DNA identical I have put in a notional father and called him Toptree.
6: Although Richard still has the "brother" I gave him from the other branch, Toptree now stands in splendid isolation and they have a new father called Ackley Williams of Connecticut born in 1789 and died 1874. This gives the direct line back to the Home Person.
7: I am now going to generate a gedcom from PAF5 and see what happens. The process will take me about three hours I think.