It should not be causing any severe problems with the game. It’s not super good, but still it’s not ancient. I managed to run M+ on a 10W Pentium J5005 and GT1030 for the memes.
If such thing happens something is not right. Run userbenchmark before running wow and check if something has low percentile. Then when the game start to lag - tab out and run userbenchmark while the game is running. Percentiles will drop but if there is a thermal or power problem with CPU or other component the percentile will drop down a lot. Aside of that you can run LatencyMon and check if suddenly system gets latency problems (due to some driver or OS misbehaving).
And you can also use HwInfo to log clocks and load (and temps) on GPU, CPU, RAM, even I/O on SSD. If the CPU is the bottleneck it will have maxed load on one core and nearly max on every other as WoW scales on cores - more if it’s CPU starved.
When I limited i5-9400F from 6 to 4 cores it did had a FPS drop in Dazar’Alor but not in old zone like Stonard and not when doing mass combat test in Karazan raid. Only at 2C config was where more benchmark scenarios showed lower performance.
That FPS isn’t that bad but if you see stuttering and lack of fluidity and it’s only you in the group that has it (sometimes it’s server side and affects everyone) then it can be some bottleneck in the system but usually not as basic as “moar CPU” or “moar GPU”.
In some cases of raid trash or encounters multiple actors cast same thing as same time with same visual effect or other effects which can cause temporary drops on weaker systems (wherever it’s some bandwidth limitations or less so - lack of combined CPU/GPU computing power).
He isn’t using a Ryzen CPU and the 50% is big exaggeration. For Intel you really need specific scenarios for it to have a measurable effect. For Ryzen it does matter but often at 2133 versus 3200MHz it can be few percent as a game isn’t solely relying on inter-core communications or large memory throughput (and likely the system running RAM sticks at lower frequency than intended XMP profile will boot them with better timings which will blur the comparison a bit). It’s bad/pointless to run the RAM frequency low but still it won’t kill the performance (unless it’s a very modern game that needs many cores and causes lots of infinity fabric traffic).
I was benchmarking i5-9400F and R5 3500X recently in WoW under Windows and then general Linux comparison and there was no magic mass performance gains there. I can re-run it for various ram frequencies in WoW too if needed as RAM clocks were tested on Linux compute and few synthetic gaming benchmarks 
So you replaced CPU to a massively more capable, you replaced RAM, motherboard, CPU cooling solution, likely also GPU or just moved modern GPU to much more capable modern motherboard/CPU. With such hardware upgrades FPS will change but that’s a multi-factor change, can’t be attribute to only one thing.
