In SA white-nationalist justifications, this belief of 'equal or later arrival' is a fellow-traveler of the Myth of the Empty Land that aims to normalize racial claims, and it's just as wrong. First, it presupposes a natural unit called "South Africa" that exists in an absolute sense, so that anyone in a group stepping foot anywhere within its boundaries has 'right' to the whole thing, even though it did not come into being until 1910--and its composition was not even settled until 1906. Even if one says the Dutch arrived 'first,' nobody with a lick of sense or knowledge will suggest that the Transvaal, KZN, or the Free State were empty when Boers arrived anymore; even the Voortrekker Monument speaks otherwise there. Second, and this is where apologists for apartheid like to move goalposts when presented with evidence (and thus why you will never win your argument), there is physical evidence that cannot be explained without recourse to iron-working agripastoralist Bantu-speakers in the archaeological record (even something as old as Mitchell's The Archaeology of Southern Africa, 2002). Finally, even if we dispense with the preceding, even they admit there were people here, but those people somehow 'don't count' even though their descendants are a significant majority in many regions. It suggests also that any people who move onto a piece of territory can't ever claim it in any meaningful way unless they are white and/or get a piece of paper from a colonial (or settler, for the Republics) government. That defies logic unless one believes white people's claims are somehow special, but if that is the case then it wouldn't matter who got where when.

In regards to Bantu-speakers' antiquity, your initial time is roughly correct for a variety of sites. But there is more. I wonder who they think built Mapungubwe (11th C.), or the remarkable stone networks and manufacturing sites of the massive Bokoni Complex in Mpumalanga (16th-19th C.)? On the former, it is easy enough to find sources; the physical traces and the cattle cultures are clearly tied to the same sets of interrelated people who were part of Great Zimbabwe's extended networks later. On the latter, Bonner, Swanepoel, and Esterhuyse (Five Hundred Years Reconsidered and The Forgotten Past) have quite vividly reconstructed what we know at this early stage in the "500 Years" project (and yes, the name is a dig at CFJ Muller's dismissive treatment in his book of the same name, published in the 60s first). There are also Toutswe and the hill cattle-cultures, who have direct relationships to the Sotho-Tswana language group, and appear 'within' South Africa in the 10th century or thereabouts. That is the very latest date one might suggest, but evidence comes earlier too. [edit 2: I forgot Peter Delius's edited Mpumalanga: History and Heritage which has a little on early sites, but the most important for this purpose is right on the cover: the famous Lydenburg Heads which only have clear counterparts in other southern Bantu-speaking clusters.]

Even if one does not like that, there are four volumes of Reizen in Zuid-Afrika in de Hollandsche Tijd (published by the Linschoten-Vereeniging early this century), two of which provide evidence of encounters between travelers and Bantu-speakers that are sporadic, but then sustained. While most involve travels closer to the Cape or are brief, as early as 1752 official VOC expeditions (the Beutler mission, for which we now have a Van Riebeeck Vereeniging volume in Engels) were establishing sustained contact with abaThembu (or 'Tamboegies') in the area just over the Kei--and there's no break in evidence before then suggesting that they somehow just swept into the area in the prior 100 years if indeed it 'took much longer.' They are still in the same broad area, having only moved in response to colonial displacement. The edge of settlement was in motion, but quite slowly, because unlike trekboers who claimed 1500-3000 morgen of land per head of household, they only expanded as groups into much smaller areas they could reliably secure and cultivate. It is also a reason why the trekboers lost (or at best drew) the early wars against Xhosa-speakers from 1779 until the British first finally decided to bring a larger, better-equipped proper military force in 1811: populations from the emaXhoseni were far denser, because they weren't claiming big chunks of land for every single settler.

The really signal work on the state of knowledge is the Cambridge History of South Africa vol. 1 (2009) but virtually any work dealing with early colonial and precolonial SA contains the actual information. However, I believe you would be wasting your effort, because they will simply claim that all the evidence is misinterpreted, was misdated, or was really the product of some other unknown vanished group. During the apartheid era, one would be committing suicide to talk about Iron Age sites prior to the VOC (as in J. M. Feeley's 1970 thesis on early agriculture in Transkei--I will have to look up the SA stats) as Bantu sites, even when the evidence was painfully obvious. You had to say nothing, or find some way to parrot the NP line. Even if they accept the argument today, such people will find a way to wave it off, because they have to--admitting otherwise would validate all sorts of things in their minds, including the idea that white South Africans have no valid claims on land. This is of course not true, but they seem to believe it, and the cry of swart gevaar runs quite deep. [edit: This flawed reasoning is ingrained enough that odious foes of a better SA on the other end of the spectrum also parrot it for political gain and to make impossible promises to frustrated voters, in a bizarre form of symbiosis neither will admit; that is another brake on receptivity.] You will not dissuade them, but hopefully some of this will be of personal interest for you.

[edit 2: This is the bulk of what I have in memory at moment; the journal Azania may have more that is useful or interesting about the earliest archaeology and its affinities, and of course the Book List is always somewhat helpful there. I will not again be in my office until Monday.]