I'm not sure I follow what you mean by 'leakage'. Do you mean noise?
If so, the minds behind the Ethernet spec thought of this. That's why the cables use twisted pairs of wires and the physical connections are transformer coupled. Additionally, signalling is differential. This is all to reduce noise and ensure error-free data transmission.
Clock phase noise (i.e. jitter) will cause bit errors if at a high enough level. Fortunately, you can check for errors by using the 'netstat -e' command on Windows or the 'netstat -i' command on macOS. On my Macbook Pro, it currently shows zero errors out of ~3.8 M frames received.
I'd like to believe you, but I'm afraid it just doesn't make any sense. If a PSU could improve the performance of a switch, don't you think the Cisco Systems of the world would advertise PSU upgrades for all their switches to get the 'best performance' and sell more product? I'm sure they wouldn't mind the additional revenue.
Here's a test you should try if your family is game - have someone switch out (see what I did there?) your JCAT one day without you knowing. Let them keep track of whether or not you heard a difference over ten trials. Could be fun.