We have to be cautious of claims of high fan speed, because of fan backpressure effects. The engineers may think the fan is very fast, say 400 linear feet per minute (LFM), whereas in reality, it could be only half of that, say 200 LFM. Or the engineer thinks it is 200 LFM, whereas it really is only 100 LFM. And so on.
Generally, we can assume that with an actual 100 LFM, the junction temperature falls by 5°C compared to the no forced-air case. With an actual 200 LFM, we can assume 8°C advantage. Admittedly, these are conservative estimates. Temperatures may actually be lower if there is air turbulence. But since we can’t depend on that always, the above guidelines are appropriate for first (worst-case) estimates.