Thanks for your outline of your approach.
My thoughts of the situation are a bit different.
I see the value of genealogy in answering three questions, which are all three inexplicably linked in specific situations:
1) How do we know (or can corroborate) that x is child of y and z (and vice-versa - ch of y & z)?
2) How do we know x is spouse of y - and when (and which wife husband is mom of which children)?
3) How do we know person A in a record at one point in time is the same person in a different record at either the same or different place at a different point in time?
The evaluation of the answers to these questions can take place in two different "universes" or "parallel tracks".
One of these universes is the universe of the "historical paper trail". It leads to a conclusion about one of the above questions as maybe/possible to probable/likely to almost certainly and sometimes a bit over and above "almost certainly". This final evaluation can be the result of the intertwining of various facts (places where these facts can be found are called "sources") and arriving at a deduced "final answer", or Non-answer, as the case may be.
This universe has one severe limitation - the identity of the father comes from the mother. It is accepted to be her husband if married or accepted as whoever she says it is if she is not. Heaven knows how many bad links we have in our files because some 3xg grandmother had an affair. We can never be absolutely sure about the father.
The same can also often be said when a man has had children bearing his name from more than one wife. Answering the questions of which wife mothered which child is sometimes a never-ending question.
Enter DNA analysis.
Another universe is the universe of DNA analysis. One use is number 3 above "how can this man be the man whose blood was found on that weapon" is used in criminal forensics. That same technology can be used in a genealogical setting (ie genealogical forensics) to reinforce, prove, or refute a conclusion arrived at from the paper trail. There are all kinds of DNA tests and all kinds of conclusions that can come from them - the details of which aren't really pertinent here.
I am not saying that assigning values to sources is unimportant or has no value, just that I am apprehensive at how much emphasis is placed on the value of individual sources and none on the value an individual source has in an argument to answer questions 1-3 above.
Anyway, back to the original premise of this thread: FTM and most genealogical software provides weak to no structure to document the sources and reasoning that are used in arriving at answering the questions 1-3 above. I appreciate your input on how this capability is already in the Gedcom standard. Unfortunately, most genealogical software that I am aware of has simply ignored it. I have concocted a way using Custom Facts in FTM to attempt to overcome that deficiency in the program in a somewhat crude way. That attempt is often just an outline of facts which has to be supplemented by a wordy narrative somewhere else in my notes. I have been submitting enhancement requests to them for years to incorporate such a structure.
RANT ON: I have come to believe that FTM is primarily concerned with keeping FTM sufficiently "dumbed down" so as to keep getting new people into the hobby and eventually subscribing to their databases and are not concerned with such complications as this or incorporating historical names into a structure of place names. It will be interesting to see if the new ownership of ancestry will make a difference. I doubt it. The subscription database is the heart of its business model - FTM is one gateway to that cash cow. RANT OFF