I didn't cite any case. My final sentence was "these are not typical numbers for such equipment, but illustrate why it does matter". So if you understand illustrate and that it was showing why all response issues combine regardless of the source you might have gotten it.
Bottom line, tubes are far, far more likely to alter the response of what they are hooked up to play with than solid state. If you like the result fine. You could get much the same with a resistor in series for the output impedance bump, and a little EQ.
What you find with many typical tube amps is flat enough response into a resistor at low power. The treble tends to roll some with power increases or distortion increases or both. Many tube amps spec some nice looking response without specifying power. Some do specify the power and quote two responses. Right off the top of my head Conrad Johnson amps often have something like 20-20khz at 1 watt or 10 watts +/- .3 db. Then full rated power at something like 1.5% thd and bandwidth of 30-15 khz. That is fairly typical conventional push-pull transformer coupled tube power amp behavior. Maybe CJ amps are not well designed.
CJ amps are well-designed, their original designers chose that sound because that is what they prefer. And they have many followers. I do have two all tube amps that I use with my HE-6, push-pull transformer coupled, an Audio Research D70 Mk II and a Copland CTA 504.