The article doesn't say which sweeteners did they study and this is a very bad style. Fortunately it links to an article on EurekAlert which in its turn links to the actual scientific paper (doi.org/10.3390/molecules23102454). It says aspartame, sucralose, saccharine, neotame, advantame and acesulfame potassium-k (ace-k) (BTW at least saccharine and ace-k are already known to be carcinogenic AFAIK) were tested agains…
I'll have to look into lactulose and palatinose.... And it looks like lactulose is out for the diarrhea reason, and isomaltulose--Palatinose is a trademarked name for it--looks practically identical to slow-motion sucrose, including caloric content. It has a lower glycemic index, but is half as sweet as sucrose, so you may end up consuming more calories for the same amount of sweetness.
Erythritol is not good for gut bacteria. It is 90% absorbed before reaching the colon, and excreted mostly intact in urine and feces, and the early absorption is why it doesn't have the diarrheal effect common to most sugar alcohols. Sweeteners that are neither absorbed nor digested draw water out of the digestive tract as they pass. Those that can be fermented by colon bacteria also tend to produce gases, that help push the water out. Xylitol is not fermentable, so the diarrhea you might get won't be quite so bad.
I think I'll stick with a mix of erthyritol and stevia, with proportions titrated such that equal-volume measures of erythritol/stevia and sucrose crystals have roughly equivalent sweetness. The flavor profile isn't quite right, but it's good enough for me until something better comes along.
(edit) I'll also add that it seems to work best in foods where other flavors can dominate. While you can sort of taste it as not being real sugar in lemonade, in sugar-free fudge, all you can taste is the sweetness. But it also forms crystals during cooling, and I haven't quite worked that out yet. Gritty/crunchy fudge is not exactly ideal.