I've got to say I agree that it would be better if everyone makes a big effort to remain completely civil and polite, this is a technical discussion really and its quite easy for these forum threads to become quite heated. When that happens any chance of decent technical discussion goes out of the window - its hard to be scientific if someone is being strongly critical of you IMO. Since we have Richard and Ehsan, Randy and Craig involved we don't want to waste the chance for good technical discussion between these guys to get a positive conclusion to the discussion.
Anyway I don't want to contribute too much to the noise on the thread so will try to keep my comments small. Ehsan, thanks for your offer to go and visit you in your lab, I would certainly like to take you up on that sometime I am next in the area, I travel in your general area from time to time so will try to factor that into my next trip. Of course that would only be for interest rather than in any way to 'check up' on trition, I don't think I could do anything of much use in that way, but it would be very interesting from a personal perspective to see your tank and lab. I am certainly convinced that triton testing is of great use even without the visit though, and I hope I have not made you think I am criticising triton.
BTW sorry for complicated english, I always write too many words
I now understand that knowing and explaining the limitations of testing will be quite difficult, as you explained, but I think it will be a good thing to work towards. IE it is worth working on a comprehensive document which explains limitations so that no one is surprised if an article like Richards shows some elements slightly incorrect. That will also be useful for people like me and others who want to know exact element quantities. We will know error margins that way which is good. You could start slowly and just add information as you find it.
As far as the hamburger thing, I understand you are saying 'triton works because here is a successful tank run with triton testing'. Thats true, but some people might be interested in using triton testing to observe other things. Maybe someone has a screw plated with cadmium in the water for instance. In your tank you don't have any cadmium plated things so cadmium is not important to you, but to someone else it might be important. So thats why I think knowing exactly what triton will report is important, rather than just knowing some tanks do well with triton. Maybe this is not tritons intention, to help with this kind of thing, IE maybe triton is only selling their service for the use of triton method tanks, but many of us are not running triton systems but still using triton testing to discover element concentrations.
Regarding certification/accreditation, I doubt its worth the extra expense if it would put the price of triton testing up. I think the service is fine as it is, but with a bit of independent testing by third parties like Richard, that will help to keep triton 'on their toes' as we say. IE it will keep triton quality high. In the same way that we like to see people testing salt mixes with triton, or rock samples with triton etc. We also like to see people testing triton themselves. Its just good practice, nothing against triton themselves.
Regarding acidification of samples, as I said in a previous comment, I personally would very much like to see triton accepting acidified samples. If its not possible then thats OK but I would prefer to acidify my samples if it will gain me greater accuracy and I imagine quite a few others would do this too, since 68% nitric acid is not essential (lower concentration could be used if the dilution was labelled and triton could adjust numbers based on sample dilution).
As Richard said, the results are not bad for triton, they seem pretty good, and they have certainly helped me trust triton results more than before the article came out. I think the best response from triton would be that the errors are in fact small, but that they will do their best to document when these type of errors creep in, and also ideally that they start accepting acidified samples, labelled if necessary.