The science

Each of the 10 enzyme molecules is a three-state Markov machine: free E, substrate complex ES, inhibitor complex EI. A free enzyme meets substrate at propensity k_on [S] and a competitive inhibitor at k_on [I]; ES either releases the substrate (k_off) or converts it to product (k_cat = V_max / 10); EI releases at k_on K_i. The off-rate is chosen so that the Michaelis constant is K_m = (k_off + k_cat)/k_on. Steps use the exact exponential waiting probability 1 - exp(-a dt) (Gillespie-style), and free molecules diffuse by dx = sqrt(2 D dt) N(0,1) for display only.

The steady state of this chain gives the Michaelis-Menten law, v = V_max [S] / (K_m,app + [S]) with K_m,app = K_m (1 + [I]/K_i): a competitive inhibitor raises the apparent K_m and leaves V_max unchanged. The same equations drive the 3D twin, which integrates the full mass-action ODE (concentrations there are in micromolar, here in millimolar-scaled model units).

Simplifications: substrate and inhibitor are held at the set concentration (chemostat), the enzymes never see depletion or product inhibition, the reaction is irreversible, and diffusion is cosmetic (the reaction is well mixed). The measured rate is noisy because only 10 molecules are counted over a 12 s window; the numerical check removes that noise by running a long headless simulation.