Giant pulses and giant micropulses from pulsars are distinguished from normal pulsed emission by their large fluxes, rarity, approximately power-law distribution of fluxes and, typically, occurrence in restricted phase windows. Here existing observations of flux distributions are manipulated into a common format and interpreted in terms of theories for wave growth in inhomogeneous media, with the aim of constraining the emission mechanism and source physics for giant pulses and micropulses. Giant micropulses near 2 GHz (PSRs B8033-45 and B1706-44) and 0.4 GHz (PSR B0950+08) have indices $α = 6.5 ± 0.7$ for the probability distribution $P(E)$ of the electric field $E$, with $P(E) ∝ E^{-α}$. Giant pulses (PSRs B0531+24, B1937+214, and B1821-24) have $α$ ranging from $4.6 ± 0.2$ to $9 ± 2$, possibly increasing with frequency. These are similar enough to regard giant micropulses and pulses as a single phenomenon with a common physical explanation. The power-law functional form and values of $α$ observed are consistent with predictions for nonlinear wave collapse, but inconsistent with known self-organized critical systems, nonlinear decay processes, and elementary burst theory. While relativistic beaming may be important, its statistics are yet to be predicted theoretically and collapse is currently the favored interpretation. Other possibilities remain, including stochastic growth theory (consistent with normal pulse emission) and, less plausibly, refractive lensing. Unresolved issues remain for all four interpretations and suggestions for further work are given. The differences between normal and giant pulse emission suggest they have distinct source regions and emission processes. Comment: 19 pages (double-spaced, draft mode), 2 tables, zero figures