I am of two minds when it comes to making sense that different devices sound different in the same circuit. On the one hand, if you take an identical analog preamplifier stage and substitute FETs for bipolars on the input it'll sound different, each and every time. To me that's not surprising, despite the fact they measure identically in circuit, but to many that's a head scratcher.
Why would that be? Within the relatively low gain, frequency and step response requirements of an audio circuit, simply swapping the input transistor types shouldn't make much difference. But the evidence of listening would suggest otherwise.
We could also ask the question in reverse: how could it be anything but different? Despite the easy demands of audio signals, passing through entirely different technologies like those of a FET or a Bipolar, it would be a head scratcher if it DID sound identical.
There are companies whose products purport to take advantage of these sonic differences, like a tube buffer between products, but I don't classify them as the same thing. In the case of a tube buffer, you're placing yet another element in the signal chain, which we all understand changes the dynamic. But when we have a simple amplifier circuit, even one with but a single device in the chain, it is fascinating that one can plug in different types of devices and get significantly different sound. Sound I would characterize as softer and less detailed for FETs and the opposite for bipolars. But, as with everything in this field, that certainly isn't always true. It depends on the design of everything else.
Most of our pure analog designs over the years have featured a combination of devices, hand selected through the listening process to give us what we are looking for.
It is, still, a craft and an art to wring the most out of our music.