This is at heart the historiographical battle between JD Omer-Cooper, Julian Cobbing, Betsy Eldredge, Carolyn Hamilton, Norman Etherington, and others in the late 80s and early 90s. A good collection is the invited book The Mfecane Aftermath that includes a lot of these discussions. There is still disagreement; I am an Etheringtonian/Eldredgian (?) myself so I have a stance based on my work in Venda and the eastern Transvaal, as well as the history of the Thembu kingdom.

Basically, the idea of murderous Zulu depopulation served two purposes: it conveniently made "empty land" for those who might claim it, or defeat Zulu claims of conquest. It also gave the amaZulu a huge amount of cachet in the tribalizing world of segregationists (1910 to 1948) SA. The very term is a neologism that confers extended unity on a very complex process.

In reality, the situation Norman Etherington describes in his 2001 The Great Treks and the Transformation of Southern Africa 1815-1854 is likeliest: the highveld always had low population densities and seasonal movement, as it does today where artificial means don't support cities. Zulu mass killing didn't normally happen; the goal was to incorporate client people. Zulu killing of independent leadership was much more common, so opponents also promoted the idea of utter annihilation to maintain cohesion. It is absolutely true that the early 1800s was a time of centralization and inventive defensive modernization in state craft (see Lesotho here) which eliminated a lot of identity clusters over time and promoted concentrated formations over smaller towns in African societies. There was conflict, it was novel and dislocation, but it wasn't the slaughter of millions once propounded. For a good explanation case of how such narratives served settlers and local leadership alike in the 19th century, Norman's 2004 article "Tempest in a Teapot?" from the Journal of African History is methodologically and narrative ly interesting.

So it wasn't necessary for mass death to clear some lands and, more often, erase the perceived legitimacy of claims on others if people were in motion. In my own work I've found that the great polities moved very little, and it seems to line up with Etherington (and John Wright) where they talk about the groups in motion that fought at Dithakong and Mmbolompo.

The question of European motive force is why I'm not a Cobbing aficionado. Eldredge really took apart his claims about slaving from Delagoa Bay as anachronistic, and even insufficient if they weren't. While pressure on land from the Cape was important, the Mpondo kingdom didn't shift, so it's not like they threw Nguni-soeakers back upon the lands of the major federations and somehow leapt over the entire Libode/Bizana region. Horseback Griqua, Korana, and Rolong riflemen in the Northern Cape and nearby were also more important to instability but for reasons of cattle raiding, not slaughter or slaving, because the Company and later the British barred the enslavement of 'Cape natives' construed broadly. Europeans (well, Boers) generated instability later as they moved into areas thought open (the Great Trek in classic pioneer-SA history), in kind of a symbiosis with the classic 'mfecane' realignments as almost chiefdoms themselves.

So TL;DR yes the era created instability and movement, but it's a bigger process that was not nearly as bloody as the mythology suggests. It was generated from within various societies and emerged through their interactions. People always had conflict and new, innovative state formations showed they were not self destructing or hopelessly lost.

Of course, others still differ. Hamilton's book is old now but still a good intro to various positions, and I'd recommend reading Etherington even if you ultimately don't share my assessment. The problem is that there is too much we do not and perhaps can never know, but that's part of the fascination too.