Or is it more a question of 'intersubjectivity' and consensus among the team members?

Do you accept the classical view of categorization, ie. that categories are based on shared properties and are defined in terms of necessary and sufficient conditions? From this view follows that a band either does or does not belong in this or that sub-genre. 

Or do you use the proto-type, or 'goodness-of-example', theory of categorization, where team members for instance could rate albums on a scale from 1-7, where 1 is not a good example and 7 is very good example of [genre]?. So that instead of an either/or question, it becomes a question of central/peripheral.

How the sub-genres accepted on PA came into being in the first place, is a different question. Some of them seem extremely internal and somewhat absurd from an outsider's point of view. Eclectic prog?  

My humble suggestion would include much fewer sub-genres, adaptation of the proto-type theory of categorization and an album-based approach. To categorize a band's entire output using a single label is not a very good idea.

Serioulsy, it depends on the artist/band - there are no hard anf fast rules and there is no right or wrong answer.

Paravion wrote: Do you accept the classical view of categorization, ie. that categories are based on shared properties and are defined in terms of necessary and sufficient conditions? From this view follows that a band either does or does not belong in this or that sub-genre.

Mostly yes, but some genres are not based in shared properties.

Paravion wrote: Or do you use the proto-type, or 'goodness-of-example', theory of categorization, where team members for instance could rate albums on a scale from 1-7, where 1 is not a good example and 7 is very good example of [genre]?. So that instead of an either/or question, it becomes a question of central/peripheral.

Not empirically - of course with artists that jumped genres we tend to go for the genre they are best known by, rather than the one they did the best albums in.

Paravion wrote: How the sub-genres accepted on PA came into being in the first place, is a different question. Some of them seem extremely internal and somewhat absurd from an outsider's point of view. Eclectic prog?

Most of the genres were pre-existing, though may have been known by different names in different countries. Ecelctic, Heavy and Crossover are pure inventions - previously they were all in one subgenre called Art Rock, but Art Rock is a broad genre that includeds artists that are not Progressive Rock.

Paravion wrote: My humble suggestion would include much fewer sub-genres, adaptation of the proto-type theory of categorization and an album-based approach. To categorize a band's entire output using a single label is not a very good idea.

What we genre we call the band doesn't change their music or how much pleasure you get from listening to them - genres and classification are a way for you to find similar music - having fewer genres would make that a lot harder.

@Dean.

Thank you for answering my somewhat polemic and rhetorical questions. 

There is of course a wide range of things to consider. One of them, as you point out, is that this site first and foremost should be helpful for people in order to discover similar music. I may have overseen that consideration, because I don't really use PA to discover new music anymore. In discovering new music, I tend to go for diversity more than similarity and care as little as possible about genres and classification because such attemps often don't match my 'mind-internal' categories.  

My 'concern' is more centered around the ontological status of categories. A lofty goal and hope for PA, is that the categories (=sub-genres) to a much larger extend reflects the way mind actually categorizes more than establishing artificially construed categories on the (lose and random) basis of the classical view of categorisation. 

The classical view is not based on empirical studies, but on the basis of a priori speculation. Since the pioneering work of psychologist Eleanor Rosch, this view has been seriously challenged. Empirical studies, determined to find out how the mind actually categorizes, show that categories are not established on the basis of shared properties and they are not defined by necessary and sufficient conditions. The proto-type theory, anticipated by late Wittgenstein, developed by Eleanor Rosh and now widely accepted by cognitive scientists, eg. George Lakoff, show that categories are centred around proto-types ( a central member of a category) and that category membership is very complex and is conditioned by individuals' conceptual system, that is in fact different among individuals. In order to describe categories, we need a notion of central/peripheral, membership gradience, family resemblance etc, etc.