I would say that if your clients don’t complain then why shell out extra money?
The only difference it could make is to allow you to mix differently, some might say better but then that depends on your current mixes and what you’re listeners are using to listen. In terms of audio quality, as you’re using imported samples, mixing in the box and exporting direct to files, the quality of the audio monitoring can only effects your mixing decisions.
However, the latency you choose can have a difference on how well the automation works, and that can depend on the audio hardware available. This is due to the fact that automation is not sample accurate (I guess it may be now in Cb6??) and it moves about more or less depending on latency. This can cause problems if you’re automating a rhythmical chopping effect for example because it’ll not quite be in time.