I am convinced that the real genesis of XBRL can be found in the financial scandals of the early 2000s (and in... hushed silence.... SOX 408).
I believe the SEC has mandated XBRL because they need the data, and that provision of that data to the investor community, while part of their mandate, is and was secondary. It is difficult enough to perform analysis of companies with semi-manual processes, no matter how much data you have. To improve efficiency, you need to import the data, and run it through your analytics before any human sees the data. The FDIC has demonstrated this with their XBRL based Call Report upgrade in 2006 (but that is a different story).
The big difference is that the SEC is not (we all hope and assume) making buy and sell decisions based on the data. they are doing what they are mandated to do: protecting the capital markets. That does not require split-second buy-sell decisions, but careful analysis, identification of outliers, and dare I say it, wondering why companies choose to not be comparable.
So getting back to the main point - the scandals resulted in that wonderful paragon of legislation: Sarbanes-Oxley, also known as the "Full Employment in the Accounting Profession Act". Of course, the world focused on the now (in)famous Section 404. And, after years of documentation, testing and (the dreaded Section 302) certifications, oh, and probably billions of dollars in auditor and consultant fees, internal controls actually are better.
But look a little further. Look at Section 408 (kindly extracted here by The University of Cincinnati College of Law: separate link).
c. Minimum Review Period. In no event shall an issuer required to file reports under section 13(a) or 15(d) of the Securities Exchange Act of 1934 be reviewed under this section less frequently than once every 3 years.
Of course, there's more. Paragraphs "a" and "b" increase the scope to include "Regular and Systematic Review" of a large number of filers.
Basically, Section 408 requires the SEC to review all filers not less than once every 3 years. It also requires that all companies meeting a set of conditions be reviewed every year (a set of conditions that I think would probably cover close to 10% of filers).
To do that, the SEC's manual (or near manual) systems would have had a difficult time coping. Section 408 in effect is the source mandate for XBRL, a mandate on the SEC that requires a data standard that can provide the SEC with computer to computer data for 'easy' consumption by computer systems. Those systems can then perform the pre-processing and analytics, flag up the outliers, and even identify some of those that should be reviewed every year. That, I believe, is the primary original driver for the adoption of XBRL by the SEC.
Protecting the capital markets by rapid and frequent review of filers, enabled by computer to computer provision and consumption of data. Apply that argument, and the requirement for XBRL becomes self justifying. All other additional uses of XBRL become an added value to the economy and investors.
The next question is: What does this tell us about the SEC's probable next steps?
First, get use to XBRL, its not going away. Second, expect the SEC to accommodate filers enough to make provision of XBRL as 'easy' as possible. After all, if the SEC needs the data for their purposes, then if concurrent provision to the markets as critical a requirement.
Finally, if the program is critical to SEC success (which I believe it is) and they will work to make sure filers do not attempt to derail the program, I fully expect the SEC to defer detailed tagging for "group 3" filers. The burden will prove to be heavier than the benefits that they will already be achieving simply from the consumption of the primary financial statements. But these are guesses only...