In a telecommunications company implementing Network Function Virtualization (NFV), the team is tasked with optimizing the deployment of virtualized network functions (VNFs) across multiple data centers. They need to ensure that the latency for end-users is minimized while maximizing resource utilization. If the average latency for a traditional network setup is 50 ms and the goal is to reduce this latency by 30% through NFV, what should be the target latency for the new virtualized setup? Additionally, if the company has a total of 100 virtual machines (VMs) available for deployment, and each VNF requires 2 VMs, how many VNFs can be deployed while still meeting the latency requirement?