Incidentally, and not sure if this matters or not, but when perusing the version that's on archive.org I was noticing a fairly substantial amount of various names misspelled or somewhat erroneous information, and this on either artist, album or label names.
This can affect searches, where accuracy may be important. (Examples: Bee Gee's as opposed to Bee Gees; Kuda Records when it's Kudu, that sort of thing...)
Are you guys interested in having all of those checked, cleared up and brought to standards? (like peer review sort of thing)
Please don't get it wrong: what I am writing below is merely just a suggestion
Beyond this, and strictly from a professional standpoint, it feels to me as if all of this information is begging to be entered into a proper relational database so that it can easily be displayed in many ways. Flat files are OK, but obviously make editing and everything so incredibly time-consuming later down the line. Having a template format designed, with proper thinking of the schema, data types, unique indexing and relations can ensure that this information will be future-proof. Once this is done, the display (rendering) can be changed at will without ever having to redo anything, and will allow for a lot of extra features to become a reality, like moderated user comments, polling and the likes.
A good example of the type of features would be: Searching by artist, and having the results returned order by format, or by release (with each format as a sub-field), or by label, or searching by format even, and each of those results would have a page of its own where additional info can be entered, graphics, comments and the likes.
It would initially mean a substantial donation of coding time from a specialist in that field, but in my opinion will ultimately be necessary in order to later free the maintainer(s) from otherwise doing redundant and tedious manual work and focus on adding more content to the site instead; as well, to basically ensure that this information survives and evolves as the display will now be completely separated from the way its data itself is being stored. These sort of sites used to cost a fair amount to build, but are so commonplace nowadays that any web coding wizard can take existing templates and easily modify them to fit the project, using a free, open-source content management system such as Drupal or Joomla.
(whether it's someone building a site of cooking recipes and ordering them by type of recipe, author, year, and ingredients or someone listing auto parts for specific models of cars, type of use, and the years of their make the problem is exactly the same viewed from the eyes of a database designer: Break down the data into simple fields and define the relations that govern them so that in turn that data can be displayed in countless ways depending on what the users desire)
In this case, the primary data types would be:
• artist field
• album field
• song field (note: Discogs doesn't have a song field, only album)
• version field (note: Discogs doesn't have a version field)
• physical format field
• release type field
• catalog number field
• label field
• year field
• musical genre field
• album comment field
• album graphic field
• artist graphic field
• label graphic field
• external link field
• artist website link field
• review link field
• popularity field (number of views)
• user rating field (registered users only)
• gateway to online store field (joining affiliate program could make the site easy commission money to defray hosting expenses)
From the start, such a site would provide its users with a variety of display formats, which could mirror the way it is now, as well as ways to arrange it by artist, by year, by format, by label, and immediately view all relevant content with automatically embedded hyperlinks in those multiple display modes. Anyone could view 'what's the most popular album in a certain category' or those types of more rewarding searches where suddenly all of this information can be leveraged and aggregated.
Later, sites designed in this way can have plug-ins that allow users to automatically 'Like' a particular page on social networks, or mention it on Reddit, Twitter, etc... which would help generate traffic and visibility, as well as become optimized for SEO (search engine optimization) which guarantees better results and more traffic from Google searches.