For defining "waterway", I would probably defer to the parameters used currently by the state to report mileage.
Though seemingly a static definition, it is not. For example, the mileage of streams in the state changes over time.
Over the course of field investigations, which may only occur once every 10 years, the length of a stream can easily vary, since the parameters to meet the state's definition may or may not be met depending on the hydrologic cycle.
The stream bed of a small tributary to a larger flow may be running a quarter mile longer or more during wet periods than during dry periods.
There's also the aquifer withdrawals from new vacation properties and permanent residences that may have been installed since the previous survey, and those diminish flows.
The test for determining stream length, as used by DEP, might be a good place to start for determining waterway.
But I do understand the problem.
In the Lebanon area local leaders decided that, as a result of Alcoa closing (and its formerly large block of surrounding properties that were previously all agricultural/woodland/wetland) it would be appropriate for zoning for the freed-up property to create an industrial park.
It sits on the main, mostly subterranean flow of the Quittapahilla.
In my mind that's foolishness.
But as a side benefit, after the stream re-emerges, a visible wetland (that was created from long previous development and road construction) was enhanced and is now an educational facility. If the new industries pollute the wetland should show it.
So, I agree it's a very difficult problem to nail down to a static, easily argued black & white issue for legal challenge and bureaucratic/government management.
But some day, we will have to face the question closely and turn a whiter shade of pale ...
![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)