... this makes sense at some wacky level.

For me, the semantic sense is that in a 0-based array, the length of the array can be used to represent its highest index. A D example:

import std.stdio;

void main () {

    auto ra = [ 'a', 'b', 'c', 'd', 'e', ];

    writefln("length of ra is %s", ra.length);
    writeln();

    write("A:  ");
    foreach (i;  0 .. ra.length) {
        writef("'%s' @ index %d  ", ra[i], i);
        }
    writeln();

    write("B:  ");
    foreach (i, e;  ra[ 0 .. $ ]) {
        writef("ra[%d] == '%s'  ", i, e);
        }
    writeln();

    write("C:  ");
    foreach (i, e;  ra) {
        writef("ra[%d] == '%s'  ", i, e);
        }

    }
[download]

Output:

C:\@Work\DMD2\posts\davido>    python_to_d_1
length of ra is 5

A:  'a' @ index 0  'b' @ index 1  'c' @ index 2  'd' @ index 3  'e' @ 
+index 4
B:  ra[0] == 'a'  ra[1] == 'b'  ra[2] == 'c'  ra[3] == 'd'  ra[4] == '
+e'
C:  ra[0] == 'a'  ra[1] == 'b'  ra[2] == 'c'  ra[3] == 'd'  ra[4] == '
+e'
[download]

(The expression  ra[ 0 .. $ ] is shorthand for  ra[ 0 .. ra.length ])