perf(core): preallocate stat responses and reduce branching (#127)

* perf(core): preallocate arrays for  responses

* refactor(core): remove switch cases for summary stat branches