Results could be expected to vary with overall water quality. Dissolved solids, combined chlorine residuals, etc. Background noise. Also, varying with ORP probe maintenance practice, cleaning and calibration. I'd suggest a trial, comparing ORP with results obtained from an alternate (eg: colorimetric) procedure. And let us know your results?
I'm not exaclty directly answering your question, but I'd bet that ORP could at least show trends, particularly in clean (eg: potable) waters, that could be checked against routine Operator data collection using a spot check / lab test. High levels of combined chlorine, such as in waste waters or cooling tower waters, etc, may increase ORP reading even with no free chlorine (just below breakpoint chlorination point, if ammonia compounds are present)? You'd probably at least increase likley hood of catching process upsets with the on-line probe.