Ships and airplanes traditionally navigate using latitude/longitude measurements. Since that coordinate system is based on degrees/minutes/seconds of arc, a length unit that uses the same system is more natural.
;-) I kind of knew this already, but thanks for confirming. Point being, there's not a good reason "why" today. Truth is that it's convention at this point. People used to use maps and various tools to do navigation by sea and air. The methods they developed assumed certain things (traveling in an arc, usage of maps with lat/long lines, etc..), which meant the knot was more preferable.
Valid points all but we're using 20/20 hindsight to render judgments on received conventions during a time in which the more logical metric system exists. A quick glance at 'Knot (unit)' in Wikipedia supplies a 'rationalization' for the use of knots involving old Mercator projection navigation maps. But all of these 'knots' and 'mile' measures are based on the English 'foot'; where did the old 'foot' standard come from? Some 10/11 convention imposed by an English King Henry III on an older system. And by extension, why retain 24 hours in a solar day or the Babylonian 360 degrees in a circle? Or more recently, why the Cartesian convention of ordinate-vertical and abscissa-horizontal on graphs rather than the other way around (sundial-clockwise would be positive; back-Kronecker [0 1,1 0] to convert)? After a while these mathematical musings start to resemble idle etymological ramblings in linguistics: eg, 'extreme' conflates 'out of' and 'sewer'; but then why those particular phonemes for the sense of 'out of' or 'sewer'? As my brother the research psychiatrist frequently retorts "You're trying to be logical. People aren't logical"
24 and 360 are really good unit bases for something you have to divide by hand. 24 gets you 2/3/4/6/8/12 while 360 gets you 2/3/4/5/6/8/10/12/15/18/20/24/30/36/45/60/72/90/120/180
And, we don't always adhere to cartesian conventions. Right-hand and left-hand rules abound in engineering. And even computer graphics regularly flips the axis directions depending upon usage (perspective transforms, for example).
There's still an expectation that a skilled navigator can work out position and movement by hand, using charts, speed and compass headings ("dead reckoning") and have those results reasonably match/confirm the information reported by more modern systems. Using knots makes this significantly easier than it would be if speed and distance were reported using statute miles or another unit that's not easily converted into change in Lat/Long.