Even though it was discussed with client the implications on having this feature enabled:

• VMs with latency sensitivity of high impact neighboring VMs on the host by reserving physical CPUs. This reduces the available CPU for other VMs, even if the reserved physical CPUs aren't being fully utilized. As a result, other VMs on the host may experience excessive CPU readiness.
• vmxnet3 interrupts are serviced by non-reserved PCPUs. As a result, if other guests on the host are consuming the balance of CPU remaining after reservations, there may not be enough CPU to service vmxnet3.

Customers insist on really having a valid use case for supporting this and insist on the feature:

We have applications that are like HPC , near real time execution is needed , for those applications we need the latency sensitivity mode set to high , this is the only way those applications manage to keep consistent performance.
Currently post installation we need to manually define the latency sensitivity mode which requires a lengthy process of taking down worker nodes one by one , defining pinned GHZ configuring memory reservation and then starting the node.If this could be done via the install config this will create a smoother installation process for us.