If we do get to 17,000, this'd make it the 6th biggest recorded volume for the Pohono gauging station. The bigger events were the floods of 1937, 1950, 1955, 1964 and 1997. Five 'big' floods in a century averages to every other decade or so. What some locals parochially call 'The Flood' of twenty years ago was only a little bit bigger than those other four were, and such floods affect far more than just Yosemite. There was no gauging station in 1862, when the Merced surely exceeded its 1997 size in a mega-flood that wrecked the whole state.
All these high water events happen December-February and are due to warm storms, which bring rain to high elevation instead of snow. There's a persistent myth that the 1997 flood was due to rain melting recent snows, but the science says that rain doesn't melt very much snow. The floods happen simply by high snowline causing tens of thousands more acres of watershed to receive rain. Especially in higher terrain with thinner soils and vast expanses of bare granite, that rain just runs off right away and swells the rivers downstream.
I share Muir's enthusiasm for such events; he witnessed and wrote about the flood of December 1871. While I shelter in place, I look forward to seeing the video record collected by the professionals.