Not sure about the maximum number and the impact of change that to a value higher than the default, but according with the following VMware KB article you can even disable that limit using the value 0 (zero), see: vSphere High Availability error messages (2146276) | VMware KB---
Senior Infrastructure Specialist
It will depend on the workload of VMs and your network connectivity.
Those default limitations are "fail-safe". So basically you will be OK with whatever (supported) configuration of VMs.
If you go higher then the IO workload of the VM kicks in. There will be large difference between enabling FT on higly utilized Exchange / SQL servers vs. some web application server for example.