The only failures we have are Aluminium and sometimes Nitrate.
The aluminium we tend to catch as it rises. Without the sample though we wouldn't catch it at all as there is a negligible effect on conductivity (please feel free to correct me). I think on average though we only see aluminium rise above the limit about 3 times a year throughout our home program which is 60+.
We have RO units where we have noticed the conductivity rise to say 20uS (incoming 85uS) - have tested the output and it still passes the levels set in the standards. Due to the rejection ratio falling we have still changed the membrane though - something I am still trying to decide if it the right decision. Does rejection ratio over-rule sample results as a safety measure???
The nitrate is only in particular locations - usually rural so I presume that it is washing from the land into the water network.
As we found out through the presentation at the 2017 National Conference a lot of these contaminants can take some time to have an effect on the body which is probably why the guidelines say a 6 monthly sample is ok.
I'm sure the chemical testing is <£20 a shot so for peace of mind we personally here will continue to do it. We tie it in with an RO disinfection anyway (which we do) so the technical time isn't wasted.