We present robust decompositions of the diffuse and discrete components of ionized hydrogen in a sample of 11 nearby spiral galaxies, spanning a wide range of star formation rate and Hubble type. Traditionally, H II region populations have been interpreted through power-law fits of the Hα luminosity function (H II LF). Here we instead compare the measured H II LF directly with predictions of population synthesis models over the entire range of completeness. In our modeling context the cluster membership function (CMF), defined as the number distribution of stars per cluster over an entire galaxy, is a principal parameter. The CMF was assumed to be a truncated power law, excluding the formation of stellar clusters initially composed of more than 106 (or 103) stars with M>1 Msolar. We conclude that (1) the best-fitting power-law slopes of the CMF vary between -1.75 and -2.00, (2) there is no evidence for an upper limit to the CMF as small as 103, and (3) there is evidence for a truncated initial mass function (at M>10 Msolar) in sparsely populated clusters. Furthermore, our H II luminosity functions do not show a glitch attributable to a categorical luminosity-dependent transition between density-bounded and radiation-bounded H II regions. This negative result casts doubt on the reality of a ``Strömgren luminosity'' feature recently reported by Beckman et al. The diffuse fraction for spiral galaxies is generally about 0.45 and shows no perceptible variation with Hubble type or star formation rate, consistent with previous studies. Finally, we have explored the consequences of varying physical resolution and sensitivity and conclude that intergalaxy comparisons are most meaningful once these have been equalized.