hey mate thanks for that information, i have just actually had a chat with VMware and they are still standing by this fact
What you are doing with that setting is guaranteeing MHZ (on a per CPU basis) in times of contention ONLY. Where there is no contention VMs take whatever they need (up to the maximum of the pCPU)
so i am confused now because thought the same as you, setting 0.26 sets the CPU max to 0.26, which would be really slow, VMware saying no, its only in time of contention and vCPU are able to use up to pCPU.
confusing.
thanks for the information though.