What is "funny" is your arbitrary rule that the root of a dim7 chord MUST lead. This happens frequently, but isn't the only scenario. Cdim7 can resolve to C in the way that Cm7 can resolve to C. This relationship is even stronger between Cdim7 and C6, where the only differences are the two interior leading notes (as one might slur melody notes). The sound is similar to a chromatic slide: B6 to C6--note the similarity to my "funny" B#dim7 explanation. Omit the "coloring" 7th and the relationship is even more apparent. You can consider the progression either as a change in quality with static root or as a "leading" slide with modified root.

Dim7 by its nature is tonally ambiguous, allowing it to resolve (or merely move) to a variety of targets. This is greatly aided by equal temperament which minimizes the disparity and fixity of enharmonic equivalents in the harmonic context, so that one pitch can be perceived as simultaneously behaving and sounding like Eb and D#. The theory becomes more complex, with one chord suggesting different roles simultaneously, but the chord naming isn't completely arbitrary, as in your "lowest note" naming scheme.