Globular clusters (GCs), the oldest stellar systems observed in the Milky Way, have for long been considered single stellar populations. As such, they provided an ideal laboratory to understand stellar dynamics and primordial star formation processes. However, during the last two decades, observations unveiled their real, complex nature. Beside their pristine stars, GCs host one or more helium enriched and possibly younger stellar populations whose formation mechanism is still unknown. Even more puzzling is the existence of GCs showing star by star iron spreads. Using detailed N-body simulations we explore the hypothesis that these anomalies in metallicity could be the result of mutual stripping and mergers between a primordial population of disc GCs. In the first paper of this series we proved, both with analytical arguments and short-term N-body simulations, that disc GCs have larger fly-by and close encounter rates with respect to halo clusters. These interactions lead to mass exchange and even mergers that form new GCs, possibly showing metallicity spreads. Here, by means of long-term direct N-body simulations, we provide predictions on the dynamical properties of GCs that underwent these processes. The comparison of our predictions with available and future observational data could provide insights on the origin of GCs and on the Milky Way build-up history as a whole.