The USGS gauge referred to (at Leonore) is about 8 miles upstream of the put-in. It lists drainage area of 1251 square miles, which is only 2-3% less drainage area than at the put-in (and former gauge site) at Lowell. Thus, (other than for the lead/lag time associated with the intervening distance), the gauge at Leonore should generally be quite accurate for determining levels in the whitewater reach described here.
The suggested minimum (500 cfs) reflects the fact that the river is quite broad, so many areas will become quite a scrape below this level.
The suggested maximum (4,000 cfs) is set only to indicate levels above the 'norm'. Somewhere around this level some features will wash out (including Wildcat, normally the biggest drop on the river!), while others just get AWESOME! Thus, many boaters will thrill to see this river 'go purple'.
Most boaters historically have referred to a gauge painted on the bridge at (above) the put-in. There are approximations of the correlation of the Lowell bridge gauge to the USGS Lenore gauge at http://www.rivers-end.org/vermilion/ and http://www.wpr.pair.com/vermilion/
There was a USGS gauge at the Lowell bridge for forty years (1931-1971). While the nature of the overall watershed most certainly has likely changed somewhat in the interim (due to increased urbanization) and climate shifts may have some effect, the following analysis from that data period should still have some relevance.
Gauge/flow analysis (based on USGS data, 1931-1971, at Lowell)
Drainage area at gauge: 1,278 sq.mi.
Minimum daily mean flow during recording period (occurred 1961.02.01-05; 1963.12.21-24): 5 cfs
90% of time flow exceeds: 11 cfs
10% of time flow exceeds: 3,050 cfs
Maximum daily mean flow during recording period (occurred 1958.07.15): 25,600 cfs
10/90 ratio ('flashy-ness'): 277 (under 3 is quite steady, over 10 is quite 'flashy')
Average annual runnable (>min) days per year: 128 (Note: some of those days may be winter, when ice would preclude safe paddling.)