SQ Of Native WAV Files vs Flac Converted to WAV on the Fly Files

I have finally got round to getting a NAS installed for primarily backing up CD rips held on my UnitiServe. Having been replaced by the Core, I might move the UnitiServe on and possibly replace it with a Core at some point. In this scenario, I would need the NAS to act as a music server in the interim, possibly installing Asset. If the NAS does a good job as a server, I might dispense with a Naim ripper/server altogether and rip the odd CD from my iMac. This could help fund another upgrade I am considering.

As most of my CD rips on the US are currently WAV files, I will need to convert these to Flac before copying them to the NAS, if I want to use the NAS as a server (because my understanding is that the metadata on WAV files gets screwed up on a NAS). The NAS (QNAP with possibly Asset UPnP server software) would then be set to convert Flac to WAV on the fly when used as a music server. I understand that once the WAV files on the UnitiServe have been converted to Flac, it is a tricky (certainly lengthy) process to convert them back to WAV. So if I go ahead with the NAS as a server, my understanding is that I will no longer be able to stream native WAV files and will have to move to streaming Flac converted to WAV on the fly from both the NAS and UnitiServe, unless the UnitiServe keeps both a WAV and a Flac version of each file which I doubt.

It is possible I have misunderstood this entire process, but assuming I am correct, my question is if anyone has detected any deterioration in SQ between streaming native WAV files vs streaming Flac converted to WAV on the fly files? If so, is the difference of any significance in your opinion? My streamer (or is it renderer) is a NDS.