"Be careful and don't live by your smartwatch -- these are consumer devices, not medical devices."
Hmm. I wouldn't "live by" any consumer appliance, but fact is, they read something, which is not described in the article or abstract and which the commentator, Mr. Fried, glosses over.
He says:
the watch measures heart rate and heart rate doesn't have that much to do with the emotion you're experiencing
Unfortunately, that's a straw man or an oversimplification. That's not the stress monitor tech on modern watches, of which I assume HR monitoring is a component. Heart rate is an entirely separate function on my Samsung Watch6. HR on my watch is also accurate (it correlates with medical devices). But HR is not being used to determine stress on my device. Reported stress levels on my Watch6 do not correlate with my heart rate. I can be at 85 bpm, working lightly, and the "stress" will go up and down. If Mr. Fried wanted to tell me something useful, he could assess the actual data being collected and the methodology employed to determine "stress." Apparently, that takes too much time for both him and whomever wrote this up.
The technique in question, if you RTFAb (Read the fsking abstract) is Ecological Momentary Assessment (EMA); the details of the data by which an EMA is performed is not clear from TFA, nor TFAb, but the upshot is it's real-time monitoring. What is very clearly described in TFAb, is that it studies one device: the Garmin VivoSmart 4 watch, whatever the hell that is, which is discontinued. What is also clear is that the abstract, in summary, says it might not be reading the right factors because it doesn't correlate with traditional methods of determining stress. Ones that take more time, data, and effort.
Not exactly surprising. I guess what we're really assessing is the quality of real-time monitoring, with a device development pace that traditional studies can't keep up with, because that device is already off the market.
Here's the TL;DR: One product, by Garmin, is (maybe) doing a half-assed job. Real-time monitoring is new, and probably not reliable when compared with a detailed panel.
The important fact is it is not "Smartwatches." It is that particular Smartwatch. What does that say about the models most people are wearing? Nothing.
This article is misinformation and clickbait. You can safely ignore it, as you have likely already figured out that you should treat any consumer device reporting health data with a base level of skepticism. You'd have to be really pretty naive not to.