The kinetic processes in nanoparticle-based catalysis are dominated by large fluctuations and spatiotemporal heterogeneities, in particular for diffusion-influenced reactions which are far from equilibrium. Here, we report results from particle-resolved reaction-diffusion simulations of steady-state bimolecular reactions catalyzed on the surface of a single, perfectly spherical nanoparticle. We study various reactant adsorption and diffusion regimes, in particular considering the crowding effects of the reaction products. Our simulations reveal that fluctuations, significant coverage cross-correlations, transient self-poisoning, related domain formation, and excluded-volume effects on the nanoparticle surface lead to a complex kinetic behavior, sensitively tuned by the balance between adsorption affinity, mixed 2D and 3D diffusion, and chemical reaction propensity. The adsorbed products are found to influence the correlations and fluctuations, depending on overall reaction speed, thereby going beyond conventional steric (e.g., Langmuir-like) product inhibition mechanisms. We summarize our findings in a state diagram depicting the nonlinear kinetic regimes by an apparent surface reaction order in dependence of the intrinsic reaction propensity and adsorption strength. Our study using a simple, perfectly spherical, and inert nanocatalyst demonstrates that spatiotemporal heterogeneities are intrinsic to the reaction-diffusion problem and not necessarily caused by any dynamical surface effects from the catalyst (e.g., dynamical surface reconstruction), as often argued.