People in the Education field have known that stress hampers learning for a long time ... but it's still nice to see empirical results.
As an interesting (and yes, a bit sad) twist on that notion - and, if I may say so, coincidentally a bit of a bridge between my comment-fellow SubiculumCode's two different areas of research - my personal empirical experience was that between ages 6-13 I had a stepfather who routinely dispensed physical punishment/discipline.
One of the harshest applications was around academic performance. If I had any bad marks, or a teacher reported sluggish work or improper behaviour, it wasn't good.
And whether as result of personal talent (ha!), or simply through being beaten and fear-driven into learned intensity, you'd better believe I was the top of my class or close to it most of the time.
But the stress part-destroyed my social abilities and integration (as opposed to academic learning), and along with other unusual childhood stressors and instability I suffered, is the biggest reason why later I spent most of my adulthood walking my own path - to the extent of being considered autistic.
Yet elite universities revel in making learning experience as stressful as possible.
People who study human memory have also known this for a long while. That is not the novelty of the finding. The novelty is the bridging of a memory paradigm of transitive inference which I believe has been shown to critically depend on hippocampal binding operations, and the effect of stress on that memory supporting hippocampal operation. I've been out of the memory field for about 10 years now moving on to autism research, but was very much exclusively in the hippocampal-dependent memory research field during my phD work. This is a good research team, and I have colleagues who have worked with some of the authors (e.g. Allison Preston). In any case, this is the type of study that is much much much more focused on current theory of a hippocampal operations supporting memory than non-hippocampal contributions (e.g. encoding / retrieval mnemonics, etc). The point is that the take home for scientists in the field won't be much like what a news clip might write about the study (but props to the submitter for giving the actual study link!)