The study of stellar-remnant black holes (BH) in dense stellar clusters is now in the spotlight, especially due to their intrinsic ability to form binary black holes (BBH) through dynamical encounters, which potentially coalesce via gravitational-wave (GW) radiation. In this work, which is a continuation from a recent study (Paper I), additional models of compact stellar clusters with initialmasses less than or similar to 10(5)M(circle dot) and also those with small fractions of primordial binaries (less than or similar to 10 per cent) are evolved for long term, applying the direct N-body approach, assuming state-of- the-art stellar-wind and remnant-formation prescriptions. That way, a substantially broader range of computed models than that in Paper I is achieved. As in Paper I, the general-relativistic BBH mergers continue to be mostly mediated by triples that are bound to the clusters rather than happen among the ejected BBHs. In fact, the number of such in situ BBH mergers, per cluster, tends to increase significantly with the introduction of a small population of primordial binaries. Despite the presence of massive primordial binaries, the merging BBHs, especially the in situ ones, are found to be exclusively dynamically assembled and hence would be spin-orbit misaligned. The BBHs typically traverse through both the LISA's and the LIGO's detection bands, being audible to both instruments. The 'dynamical heating' of the BHs keeps the electron-capture-supernova (ECS) neutron stars (NS) from effectively mass segregating and participating in exchange interactions; the dynamically active BHs would also exchange into any NS binary within less than or similar to 1 Gyr. Such young massive and open clusters have the potential to contribute to the dynamical BBH merger detection rate to a similar extent as their more massive globular-cluster counterparts.