A few more thoughts on the matter. I'm quite comfortable with using the LNAB field for the patronym; that's what has been decided by all the Scandianvian Projects here on WikiTree. The problem is that there are a few extremely common patronyms, like Andersen, Olsen, Hansdatter. So to separate one Ole Olsen from the other, some other identifier has been common. In Norway, the most common practice is to append the "Farm name", like "Ole Olsen Berg". The farm name is not really part of the name; if Ole moved to the farm Dal, he would be called "Ole Olsen Dal" in the next parish register entry. Many lived for generations on the same farm, and modern family historians tend to think of the Farm name as a kind of surrogate surname.
Now, there are two ways to handle the farm name, and both have their pros and cons. One may enter it as the "Current last name", and our Ole will show up in family Trees as Ole (Olsen) Berg, which is not so bad. But on his profile page he will show up as "Ole Berg formerly Olsen" which is positively wrong. He never gave up his patronym and at his burial he will probably be found as "Ole Olsen, 70 years old, Berg".
The other way is to enter the farm name in "Other last names", which is what I personally prefer. But then he will only be "Ole Olsen" everywhere but on his profile page, and the Berg farm name won't even be indexed.
There is a third way, which should be called an anti-pattern: Entering his LNAB as "Olsen Berg". We actually see that on thousands of old GEDCOM imports, because most genealogy programs only have one surname field. It looks nice both in the profile and in family trees, but neither Olsen nor Berg will turn up in searches or as possible matches.
My preferred solution would be to have the "Other last names" indexed and appended to first and last name everywhere. Hopefully, we might also get a report to weed out the "Olsen Bergs".