they moved the flight to ATL with a tech stop in FLL.
Not a "tech stop" as SA was able to and did sell JNB-CPT-FLL, in addition to ATL as a terminator.
SA's ATL-SID-JNB however, was.
It was a tech stop in the sense that the 744 could not fly CPT-ATL non-stop so had to stop somewhere. It did not stop in FLL on the way back and to me that says something about the type of demand that there was in S.Florida for this flight. That and the fact that AA has shown tepid interest at best in flying to SA from MIA despite having a baked-in feeder partner in SA.
Having had the experience of actually working on the commercial and operational planning, and network strategy of SA flights to the US in the 90s and early 2000s, I'd say this is mostly wrong, but with some elements of truth. Basically, you cannot put these decision in binary boxes of tech-stops, etc. Here's some context (apologies for the longish post):
When SA restarted and grew US operations in the 90s they initially started with JNB-JFK-JNB. They could fly JNB-JFK non-stop with the B744, first of which they received in 1991. However, due to hot and high challenges in JNB, they would have to take a significant penalty given that the B744 would not be able to get out at MTOW. Thus, they used SID as a tech-stop. It was well placed on the GC route and it has some historical significance. In the early 90s, tech-stops were far more common and accepted and it didn't create a significant competitive disadvantage. They did have 5th freedom right for SID, but the local traffic between JNB-SID-JNB and SID-JFK-SID was tiny. The return route from JFK went non-stop from 6 days per week and stoped at SID once a week for local traffic. This is indicative that this was as close to a true tech-stop as one could imagine, even though they had 5th freedom rights. As time moved on, SAA got more creating looking for 5th freedoms through bigger markets and operated JNB-LOS-JFK-LOS-JNB for a while. It was particularly lucrative, until the Nigerians canned it. This became a feature of SA's US routes for the future and SA operates through DKR and ACC as well in various route configurations. This combined tech-stops with commercial opportunities and advantages.
When they started MIA as a second US light, the strategy was to use MIA as the transit hub. While passengers could transit on AA at JFK, the connecting options were simply not as significant as MIA. There was certainly some O&D traffic from MIA, but that was never the purpose of off SA's strategy. JNB-MIA is a very similar distance to JNB-JFK (less than 100nm more), but the problem is that there wasn't a place to tech-stop anywhere near the GC route. SID adds nearly 200nm, and with more southerly flight paths being preferable at times, it could end up adding up to 500nm from the optimal track on a given day. The other options were WDH, WLB or CPT (and even LAD was considered). WDH and WVB would be the shortest, but not add significant value in terms of local traffic. WDH would also present some hot and hight issues (still can't get out of WDH near MTOW given that it's only marginally lower altitude than JNB and a lot hotter at times). So you might end up still taking a payload hit at WDH thus negating the purpose of the stop, however, WVB is sea level and plenty of runway, but even less local traffic. From a purely operational perspective, WLB would
CPT offered a different option as a "tech-stop". A much bigger diversion from the GC route, but had the advantage of picking up significant local traffic and not having to connect them to JNB in the first place. The critical part is that the return flight MIA-JNB went non-stop, not stoping at CPT. So CPT was a tech-stop, albeit one that was chosen for commercial rather than strictly operational reasons.
When SAA switched partnership from AA to DL, and thus MIA to ATL, the original intent was to keep the same configuration and strategy. Given that CPT was already a tech-stop, adding a second tech-stop at FLL seems unreasonable from an operational perspective (could go with the one-stop at WLB). It had a commercial benefit too in terms of local traffic, although it's arguable if that that was ultimately successful. Additionally, the flight could have easily been dispatched CPT-ATL non-stop since the B744 can get out of CPT at MTOW regularly, and it's not clear that you'd have had to leave too much behind to carry the fuel. If it was purely about operational challenges, the flight would likely have been planned as JNB-WLB-ATL, the commercial and strategic interests had other ideas. Tech-stops are not always just about the operational planning, but commercial benefit too. This all became moot when the A346 arrived two years after SAA switched MIA for ATL.