I'd go by what the results with a reference solution or standard say, not with which value is the most convenient. Numbers on salt buckets can be off. And if you have 3 tests with 3 results, that doesn't mean that the average of them is correct. They might all be off, or the most extreme one might be correct, without checking you wouldn't know.
In general I wouldn't even use a test that I haven't checked against a reference solution. The one by fauna marin covers a bunch of parameters. Some you can DIY.