We present a new simulation technique to study systems of polymers functionalized by reactive sites that bind/unbind forming reversible linkages. Functionalized polymers feature self-assembly and responsive properties that are unmatched by the systems lacking selective interactions. The scales at which the functional properties of these materials emerge are difficult to model, especially in the reversible regime where such properties result from many binding/unbinding events. This difficulty is related to large entropic barriers associated with the formation of intra-molecular loops. In this work, we present a simulation scheme that sidesteps configurational costs by dedicated Monte Carlo moves capable of binding/unbinding reactive sites in a single step. Cross-linking reactions are implemented by trial moves that reconstruct chain sections attempting, at the same time, a dimerization reaction between pairs of reactive sites. The model is parametrized by the reaction equilibrium constant of the reactive species free in solution. This quantity can be obtained by means of experiments or atomistic/quantum simulations. We use the proposed methodology to study the self-assembly of single-chain polymeric nanoparticles, starting from flexible precursors carrying regularly or randomly distributed reactive sites. We focus on understanding differences in the morphology of chain nanoparticles when linkages are reversible as compared to the well-studied case of irreversible reactions. Intriguingly, we find that the size of regularly functionalized chains, in good solvent conditions, is non-monotonous as a function of the degree of functionalization. We clarify how this result follows from excluded volume interactions and is peculiar of reversible linkages and regular functionalizations.