In the last five years, most of the work I have done has been about a handful of individuals--cases much like yours. I have a good hypothesis, seem on the right track, but no amount of logic and reasoning can bridge the gap--the records just don't connect one personality to another.
I suspect a difference between our work is that I usually don't have birth and death records, per se. Although they state the obvious, a few of the things I keep in mind follow:
- Working back in time is more difficult than working forward in time.
Excluding burned courthouses, some broken families and other anomalies, there is more historical and personally identifying information available as time passes. I start with my connection, by the time the work is finished, I want enough information so the case can be analyzed from both perspectives (future to past and past to future).
- There often is no direct evidence that connects a person later in life to their birth identity, much less to identifiable birth parents. (Or such information is not readily accessible, formally archived or indexed.)
- Each person (even each personality) can be compared to an iceberg--there is always more to the story. My preference is always to learn more about those to whom I know I'm related.
William Harley (3)
(1) You mention when William Harley (3) died and when he married, but nothing about his children. Perhaps there were none. Recognizing the possible difference in record circumstance (USA vs Australia), I often learn many clues about the parents (where they were born, for example) from the records of their children. Naming patterns? The records about the children would include the children's birth, marriage and death records; it would include the census records about the children.
(2) You know the date of death for William Harley (3), but what about his obituary and/or burial? Are there funeral home records (or even cemetery records) available? Was there a church service? Working with all the records created at the time someone died may provide more or slightly different information about someone's birth, age at death, parents an/or immigration. Surviving siblings? Military service?
(3) Somewhat along the lines of no. (1), this personality married in 1863 and died in 1912--I would expect there to be many records "about William" that were created in that time period--newspapers, voting lists, occupational notations, court records, land records ... In the course of gathering such research, do you become aware of other Harley families and/or other men William Harley of the area.
William Harley (3) and William Harley (2)
Are you able to join or break these two Australian personalities? According to Wikipedia, Moreton Bay is about 45 km from Brisbane. From the different vital records, tax records, newspaper archives, church records, apprenticeship arrangements, etc. of the area, are you able to develop information about men so named and/or other Harley families about Brisbane during this time? As presented, there is a 15 year gap between the records about (2) and (3). How can you narrow that gap?
William Harley (2) and William Harley (1)
Separate from the work about nos. 2 and 3, are you able to join or break these to English personalities? Are there more court records available about the case? Were there witnesses; was testimony given that might provide describe his circumstance at the time the charges were filed/hearing held? Any address for the accused? Newspaper items?
William Harley (1)
You have a nice collection of records about the family, but it wasn't apparent that you knew very much about any one of them, or the family as a whole. Somewhat as above, would circumstances about this William (1) or the family help to explain the ham?
Obviously death, obituary and probate records about the father and/or mother could be rich in clues (even clues that would associate no. 1 and no. 3). It is possible this information relates to William (2). Ditto above, what circumstances might help to explain the ham incident? (William wasn't the first born son and there is no reason to believe he is not able bodied.)
Separate from a "quick fix," I would be interested in the family group about William Harley (1). Lay out a basic framework--dates/location of the births of the parents and all of the children, all of their marriages (and spouses) and when/where they all died.
With that framework in place, you should be able to set some priorities and work to exhaust the available record groups. As I'm sure you know:
- Obituaries of family members sometimes carry the golden phrase, "Is survived by ..."
- Children, especially those who died young or did not marry, are sometimes buried in the same plot and/or cemetery as their parents.
Separate from the information found in record so discovered, my own work almost always benefits from the search itself. Ala, just how many men "William/Will/Bill Harley" appear indexed (or not) in those same record groups.