Decide the (short/mid-term) fate of Multicore

[Kukovec]

In service of multicore, an experiment initially conducted by @konnov , several classes (e.g. cahces) implement serialization features, which are redundant in the current implementation. We need to decide on the timeline/feasibility of multicore, since disregarding the serialization features would allow us to significantly simplify the architecture in the refactoring of these utility classes.