After three years at 10 - 20 ppm, I'd say it is just about right. If it gets higher your plants are probably in trouble and not taking up enough NO3. Below this your plants will be in trouble.
I agree, except at higher levels, this is not an issue.
The plants should not be in trouble either way if the dosing is higher or not......
in other words, some other dependent factor is causing a problem, not the NO3 themselves.
The range is likely up to about 80--100 before you see a slight decline in growth(see Gerloff 1966), but then again, hydroponic solutions like modified Hoaglands are 235ppm N and mostly NO3 since NH4 will burn roots. such solutions are what are used to measure non limiting growth as well as are used for Tropica and most aquarium plant commercial growers.
Gerloff and Kromboltz (Paul who still post every so often on forums) used a 1/5th dilution of the hoalglands to do this study. I see no issues, but they also did not use CO2, sediment based ferts etc and the study was for 8 weeks, not 2 years etc.....
Also, luxury uptake may not be that in natural systems which vary greatly in terms of nutrients. Times are plenty and times are lean, so unless you look at the entire seasonal cycles..........it may not show the whole story. Under horticultural conditions........perhaps, but not natural ones.
Still, a target of 10-30ppm is what I and the group in the Bay area cam eup with about 15 years ago and it's been pretty good, some suggest 5-10ppm, eg PMDD or the plagiarized version of PMDD: PPS.
This works, but care needs to be done to ensure the testing is accurate and the dosing is consistent, light is lower etc. No reason why that cannot be done under those conditions.
Some weedy stem plants did not do so well and there was less need to maintain such lower levels at those ranges, and many never calibrated test kits for NO3, so they often bottomed out and adding a bit more buffer room seems to help that human issue. EI was a response to that issue. Made management and the likelihood of errors easier for most folks+ required less testing/expertise.