A resource pool isn't going to make much of a difference unless the ESXi host on which the VM's are running is under contention. It sounds like your host is fine.
As the other poster asked, how are you determining that your VM's are having CPU constraints? What metric did you look at, and was it from the guest OS point of view or from VMware point of view, or something inside the application running in the VM?
Does your VM (or the application running inside the VM, for that matter) support muliple CPU's?
