TL;DR - OK where the father is known, otherwise not OK; time spent on saga genealogies better spent elsewhere.
Speaking to this in terms of general practice, patronymics are often "added" to people who are not known to have used it specifically in their lifetime. This is only OK in the very specific circumstances where it is known with certainty what the name of the father was. Thus, if you have a (primary) source that says that "Sven i Kjärr" was the son of "Anders i Kjärr", then it's perfectly fine to refer to Sven as Sven Andersson. I would say that this is fine for the historical period - for example King Erik Eriksson of Denmark is referred to as such (albeit in Latin) in 1135 ("...Fvndacio prebende canonicalis per Ericum regem Erici regis filium. In nomine sancte & indiuidue trinitatis Ericus. erici filius...", c.f. SDHK 180 where Erik Eriksson of Denmark establishes a benefice at Lund).
However, if the same documents state that "before Sven lived at Kjärr, then it was occupied by Olof..." then there's no proof that Sven is the son of Olof. It's possible but by no means certain. Consequently, "Sven i Kjärr" should not be given the patronymic "Olofsson". This can lead to a lot of down-stream assumptions: because a profile or record states that Anders' patronymic was Olofsson, then Anders must be the son of Olof i Kjärr. For which there is no evidence. In other words, never assign a patronymic to a person unless you're 100% sure of the relationship.
An example of exactly this situation came up yesterday in a discussion on G2G here where a Joan i Angestad has been assumed by the creators of no less than three profiles to have been the son of a Magnus. On closer inspection of the available material, this is by no means certain. Even more problematically, Joan had also been given an additional surname of "Blix" - with source documentation that never calls him anything but "Joan i Angestad" (on the same assumption that he was the son of a Måns Blix - family names were not hereditary at the time). The best I can think of in a situations such as this is to give him the LNAB of "Unknown" and a nickname of "i Angestad" as the farm he lived on is what defines him in the available documentation.
I must admit that I'm very confused about profiles dating from before, say, 900 CE in Scandinavia. It appears that a lot of time and effort is dedicated to making spiffy, nice profiles for people who never existed. Although there may be an historical kernel to the sagas, do recall first, that they were written no less than about 2 to 3 centuries after the events happened. Secondly, just because it was written down doesn't mean it's accurate (oral tradition or not), and assuming that books of what is largely myths and legends have any form of detailed genealogical accuracy is like asking, say, an academic historian specialising in the Plantagenet period to recount an accurate genealogy of the Anglo-Saxon families (including cadet branches). It'd be a muddled mess at best, and disaster at worst. Thirdly, none of the information can be corroborated by outside sources until we get down to the period around 1000-1100 CE, with (I think?) Adam of Bremen being one of the earliest non-Scandinavians taking an interest, and mentioning (for Sweden) Eric the Victorious and his son Olof (Eriksson) "Skötkonung." For Denmark, I would accept Gorm the Old as the "end of the line", and the Norwegian kings are such a mess that I wouldn't know where to begin, but it would, for certain, not be before 900-1000 CE in any case. With all the profiles that need lovin' wouldn't it be better to spend the time on actual people that we know existed?
Rant over! :)