Dear Andrew,
Thanks for the compliment. Enough progress on these wretched Flintshire familes for one night. There seems to be an even greater degree of intermarriage between them than is usual even in English gentry marriages, probably because North Wales was (then as now) relatively underpopulated, so there were fewer eligible marriage candidates.
I frankly admit to plagiarisation, which is what almost all genealogists do anyway except for original research on ordinary (i.e. non-gentry) members of the family immediately above them. I am lucky in somehow having access to a site (stirnet.com) where someone has put in a lot of work on linking traditional pedigrees with each other; and (which is rare) indicating uncertainty or probability where it is not clear where the exact relationships coincide. (For example, on one pedigree you may have a wife who is the daughter of XX, of Y, when there may be two or more XXs of Y and she has been left off the pedigree of the Xs, This is sometimes easy to resolve, if you have reliable dates on both sides, but often you don't). I try, though probably not consistently, to put a note in the "About" Section when there are two possible candidates. Stirnet.com can be maddening unless you are prepared to pay (which I'm not); it normally cuts you off every five seconds so it takes ages to copy in data. Like any source, it can be wrong, and it often has lots of errors of omission; its main sources are heraldic visitations, Burke's Peerage, Burke's Commoners, etc etc, and regional published genealogists such as Dugdale (for Yorkshire): Vivian for Cornwall and Devon etc. You can find a lot of these in the Harleian Society publications. if you just look at random on the internet, you frequently get rubbish endlessly rpeated, and some sources, for example, for Americans, "Our Plantagenet Ancestry" should be actively distrusted for their identifications of American immigrants with the people allegedly born in England.
As for the Salusburys, a lot of their early pedigree seems to me to be dubious because I have not seen it matched against otherfamily pedigrees. There were two other North Welsh families, the Salisburys and the Salesburys, which lead to extra possibilities of confusion; the Salesburys seem to be an off-shoot of the Salisburys, and the Salusburys may be an offshoot of the Salisburys of Denbighshire - and the spelling differences seem to be quite deliberate, from an early time, to distinguish the families, rather than (as one would expect) random spelling differences when people frequently spelled their names in different ways.
My other main source for this lot and their relations is the "History of Parliament" biographies, available on-line for almost everything after 1509. Again, like any source, it is not infallible, usually with errors of omission rather than commission and with varying degrees of detail and reliabilty between the volumes published earliest and those published more recently. But even at its least best it gives the reality check that someone relatively recently has actively researched the details of someone's life, rather than simply copying someone else. For this level of English/Welsh society you usually get a confirmation of a genealogy for maybe around 10-20% of males in a family; plus clues to others - plus the occasional pleasure of finding a 70-year-old Member of Parliament who married a young woman as a last desperate attempt to have an heir, had a son, threw a wild christening-party, and everyone got so drunk that 16 people died in the fighting that broke out. So many of the dates in genealogy before the sixteenth century (and quite a lot afterwards) are guesses that no-one would have put in the son as being born 70 years after the father.
Rootsweb can be (but is not always) very helpful. Where a pedigree exists on rootsweb and matches with the other pedigrees, you sometimes get a lot of extra information. If it's sourced with dates and places (for example, date of will, date of inquisition post mortem, place of residence) you often have information that you don't get in the printed pedigrees. But even the most reliable Rootsweb pedigree is going to run into areas where one person has simply not been able to research everyone who has ever lived, and may repeat misinformation from other sources.
Geni is still pretty well in its infancy. There are huge missing areas, and always will be for most of the English and Welsh peasantry. Even in the gentry families there are lots of uncertainties which can only be resolved by balancing probabilities or loking at wills or inquisition post mortems (which most of us cannot do on-line) or other details of individuals' lives.
So sources are everything. You can spend your life researching someone (and possibly come up with very little) or spend ten minutes adding someone who connects with somebody else on Geni. If possible, add the source: but think about its reliablity in the other connections around it. I would almost never trust a source which only gives so-and-so as having a father XX and a mother XY, with grandparents ZX and AZ, and BY and CY. Even early medieval lines of Kings are sometimes pretty doubtful. There are (for example) five normally accepted lines of descent from Charlemagne, even though he had masses of children by multiple partners, and one of these lines looks to me like a guess which may be right or wrong. The Bosonid Kings of Upper and Lower Burgundy are linked by a woman whose dates and those of her children do not seem to match up; my guess is that it is actually two women who had the same name in roughly the same area at roughly the same time, and who were wrongly mismatched a couple of centuries later. (You or I might have been baptised as Boso, by the way since it was once a common name, if it was not for the twelfth-century theologian Abelard, who wrote theses in the form of alleged discussions between someone and a "Boso" in which Boso is always the idiot who is taking the wrong side, after which the name rapidly went out of fashion).
Mark
