If you are speaking of family members of a 200 year or older profile, then that will include countless "descendants". That is why there were / are socalled Massive Merged Profiles in the first place. With all the name variations included. Endless duplication. And even disconnection of validated parent profiles because of faulty secondary sources.
In this specific project (COGH or Dutch Cape Colony project) we try and source the spelling of the LN'AB in the first place by if possible both an image of baptism and transcription thereof. This includes the parents (and the Dutch were also so prude as to include who was "the supposed" father or who was not ["onecht"]).
The "settlers" [the socalled A-generation] have roots in Europe or the far East. As far as those go, we try and "fix" the spelling of the LNAB as near as possible to the original spelling until such time as we can source the baptism or primary Dutch, German, French, Danish [etc.etc.] records properly.
Active profile managers have a vested interest to be part of this validation proces. They are eager to be part of this and help. We relinquish our individual familial claim of managerial "ownership" in favour of a project that through it's connection to a forum [Google-group] becomes the custodial guardian. Everyone can still post and edit away, but there is a monitoring system in place. We thus build the trees downwards from the first settlers, and validate them, profile for profile, generation for generation until we get to the beginning of the nineteenth century, when things become murky again with huge families on the move with multiple wifes and complex naming conventions.
This has been said over and over again - nobody loses the right to add or edit (even if pre-1700 certified or not). Only the exact spelling of the LNAB and the parental connections are protected. If needed (through collaborative discussion and consensus) the protection can be lifted and the needed changes done.
What will eventually rest is a framework of validated family lines from the start of the Dutch Cape Colony settlement in 1652 until the British invasion in 1805-1806. This way duplication will be made redundant (because the intention is also to have an index at hand with name variations - much like the Acadian project has).
This is were curation comes in.
This is for me specifically important because I have DNA "evidence" that I belong to a certain haplogroup of one specific Huguenot ancestor from the mid-1650's [just one of many South Africans with the same issue - we are all cousins]. Because of the state of the nineteenth century profiles with it's endless duplication and faulty sourcing, it is paramount to create the trees downward as validated as possible, which also implies the correct parents. This implies scholarly validation through extensive collaboration and communication.
But then South Africa [as a colony] grew from a few thousand to about 19000 at the end of the 18th century. And there were many many records kept. So it is perhaps why we South Africans and descendants of South Africans have less of an issue giving up some sovereignity [which as we all know in WikiTree does not really exist because nobody owns a profile let alone one that is 200 years old or more] in exchange for better sourced and validated profiles.