For there to be *any* difference, one of the amps must be audibly
distorting. Most SS amps don't do this,

A reasonable design, operating within the linear region and driving a
benign load perhaps. Do all transistor amplifers sound the same driving
1 ohm loudspeakers?

so the odds are that you'll be hearing valve amp
artifacts if there is an audible difference. These are typically
euphonic, i.e. 'nice' even though inaccurate, so there may well be an
expressed *preference* for the tube amp.

The study of this would be the point of the exercise.

However, this has nothing to do with fidelity to the source, where
you want *zero* audible differences.

Ideally one would want this as an option.

What was the point of your question,
and what did you mean by 'in favour of the valve amplifier'?

To compare a transistor and valve amplifier where the former is working
in as linear a manner as practicable but there is a strong preference
for the sound of the latter.