That is not far off. VM #2 will get processing time, if requested, but VM #1 being active will cause delays in VM #2 getting CPU time. When VM #2 finally gets some cycles VM #1 will pause while VM #2 uses its slices of CPU time. It is also important to note that periodically, even if VM #2 is doing nothing but is still powered on, all vCPUs of VM #2 will check in and briefly require 6 cores. This will interrupt the processing of VM #1.
This scenario will also cause very high CPU Ready time (time when a vCPU is waiting for pCPU time).