People that like tubes, me included are people that love colored sound. They measure like absolute crap dont they ? If people like cables , same thing.
It isn't the same at all.
Valve power amps alter the FR of the speaker they are connected to due to the interaction of the higher output impedance. This is a simple fact as is the fact that we are very sensitive to changes in FR. Also most output transformers have a lot of distortion at low frequencies, since most of the music volume is at low frequencies this distortion adds loads of harmonics in the mid frequencies, this may be audible too. People may well like this, and that is fine by me, but it is not "high-fidelity" to the input signal and should not be pushed as such.
Cable bollox on the other hand is demonstrably a con, there is no mechanism whatsoever by which one properly engineered cable could possibly differ from another at audio frequencies. None. Just lies and legal thievery.
I tend to buy nice stuff and keep it a long time since audio is a mature technology and not that challenging technically.
About 10 years ago fancied a DAC that could decode files bigger than 16/48, my system was Goldmund Mimesis 20, 22 and a pair of 29.4 monos. I tried several all expensive by ASR standards, the dearest being a Linn Klimax.
In carefully level matched comparison there was no audible difference between any of them, which in retrospect there wouldn't have been since all of them have distortion, noise and FR better than human ears can detect. I did hear a difference between some of the dCS filters, but they influence the FR so not unexpected.
Since then I realise when the system is sounding particularly good it is me or my mood not some new component, since there are no new components.
And I compare direct with original analog tape recordings on a Studer A80.
As an amateur recordist, starting with a valve mono reel-to-reel in the mid 1960s and graduating eventually to a Revox B77, which I still have, I know that there is no analogue recording method where the output of the recorder sounds indistinguishable from the microphone feed. An expert can get reasonable recordings but on the sort of music I record if the levels are set to make sure no noise is audible the peaks clip. Luckily tape recorders clip gracefully (a popular limiter for digital recording is a tape recorder emulator) so as long as it isn't too much it sounds nice enough.
In fact the first recorder I used where I couldn't hear an obvious difference between the recorder output and microphone feed was DAT, a StellaDAT which I also still have.
So expecting me to take the output on an expensive tape recorder as a reference rather than expensive current fashion bollox is also unconvincing based on my experience.