I do want to limit the surface of the storage network and the complexity of managing large numbers of dual-homed VMs and NFS Exports.
As I worked through this solution, we came to an agreement that a file server could be used for multi-VM file access, but the datasets were expected to exceed 2TB. My idea was to setup a file server in their development, test, and production networks, dual-home those three servers, mounting NFS volumes directly to the VMs.
Further research revealed to me the increase of VMDK limits from 2TB to 62TB in vSphere 5.5. I understand volume extents, but never felt comfortable putting production sets on them in the event of some type of corruption. With that said, my approach will now just be to build out file servers in each environment to negate the requirement of volumes being mounted.
If anyone else has insight on this, by all means...
Thanks for the response titatnium...
Greg C.