Less than desirable network cabeling

I have been working from home recently and decided that for several mental/psychological reasons it was better to turn the empty bedroom into a makeshift office versus using my main home machine.  Just to provide some separation of work and non-work.  Since there is no end insight for working at home,I thought it better to run a "hard wire" to my work machine vs. using wireless, in order to a connection with potentially higher speed and more reliable. 

The previous owner of this house had wired this place up to be a small call center.  The downstairs living room has something like eight RJ11 jacks.  The bedroom with my office has four.  There is even a "58-block" (aka a punch down block for RJ11 wiring) in the furnace room.  The cable used is a CAT3, 6 pair.   Well RJ45 is four pair....hmmm....I am not proud of it, but over the years I have converted these RJ11 jacks to RJ45, and just wrap the unused pairs around the outside of cable.  Yes, I know this violates all sorts of best practices, it really bothers my inner OCD, and I am even embarrassed by it.  However, it works, and works fairly well.  I had a chance this week to benchmark the differences.

First benchmark is the of the office machine connected via WiFi, the WAP is a 5ghz "N" that advertises up to 300Mbps, it is physically located in the next room over.  

Second benchmark is using the ghetto Cat3 cabling.

Last benchmark is using a proper Cat6 cable plugged directly into my switch.

The Loss section of the benchmark can for the most part be ignored, as it is an instance reading vs. a running average.  Although just from watching it, the loss did seem to be a bit higher on the Cat3 vs Cat6.  It would seem that at least in this test the cabling makes little difference.  That being said my servers are all directly plugged into the switch using at least Cat5e if not Cat6.