In one hand, there are names from other languages that, obviously, follow the rules of those languages (Niamh, Sean). Those names are legit in THEIR languages, they are not English names, even being used by English speakers.
The same goes for transliterated Arabic names. Even if the educated speakers know what a group of letters represents in Arabic transcriptions, that doesn't mean that the use of this group of letters is legitimate when spelling English names or words.
One thing is if those names are legitimate (yes, they are in their languages), another one is they can be considered legitimate in English because some of them have been used for a long time (it would depend from name to name) and another it is if legitimate to use foreign names or not. In this last case, it is a problem of acceptability.
Qameron doesn't follow the English accepted standard rule, the English rule; that means, Qameron can be logical, if you want, but it is not legitimate in English. If the subgroup using the spelling Qameron succeed and a critic mass of English speakers accept Qameron as a standard spelling, then the English spelling rule would vary and the name would become legitimate; until then, it is not legitimate.
The names have standard spellings. This spellings are determinated by tradition, stablished by authoritative works, as Biblical translations, for instance.
That one person use a personal spelling doesn't convert this spelling in a legitimate one, just in the same way that with the words. It is only when a critical mass of speakers, mainly educated, accept it as acceptable that it will be legitimate. Until then, it will be just an eccentric use (with historical, sociological... value).
When referring to some individual, obviously the spelling of her or his name will be respected, just as if he or she chooses to spell the name with numbers (z859d), with symbols (as The Artist Formerly Known as Prince or the urban myth L-a) or whatever. That doesn't legitimate in any way the form, because to be legitimate a name has to conform to the linguistic rules, among them spelling (and numbers and symbols are not elements usable when spelling words and names), and the accepted standards.
The editors don't "correct" spellings of names when referring to real people because that is the label which they are known, not because that label is correct or legitimate. If the name, for example, is used in a school text book, not in reference to anybody, it will be corrected and, for instance, Jhonatan will be fixed as Jonathan. The same is true when editing authors who use personal spellings in their work: the spelling is respected not because it is a legitimate spelling, but because it is a piece of art (I'm thinking specifically of Juan Ramón Jiménez in Spanish and his spelling).
Finally, in the USA the examples of groups of letters used as names that challenge the orthophonetical English rules, in the way "Mara>Sarah", are not unheard (I can check in the other board which I take part in, if necessary). If those names would not be orthographically legitimate and they are used in the USA by English speakers, that means that the concept of legitimacy is also valid for the English speaking countries.
What it would be different is the process to become legitimate (speed, diffusion, scrictness...) and the scope of the acceptability and the usability concepts.