That's the prediction from iSuppli (via MocoNews).  To get more granular, they forecast that digital will represent 40% of total revenues (up from 12% in 2006).  And they cite several trends including:

-Mobile: competition with broadband, the fact that music-enabled phones out-sell MP3 players 2:1 already, and the decreased piracy risk of the mobile platform
-Vertical integration of digital music services (think iTunes and Zune).  [Ironically this is occurring while the record industry is horizontally integrating]

While bold, the prediction on its face doesn't seem far-fetched to me.  But thinking through how this will happen shows us how far there is to go.  Will the incumbents like iTunes drive the lion's share of the growth?  Will new entrants like Amazon emerge?  There are few players that have achieved any major traction in mobile music, primarily because the user experience isn't yet there, so the growth curve for mobile will likely be steep.  And there will need to be at least 1 or 2 more generations of software/hardware to get something that is ready for the mass-market. 

Similarly the labels will need to continue to change their organizations to cater to this world.  Licensing for digital stores needs to get even more streamlined than it is.  The labels should also help their licensees obtain publishing rights more easily.  More aggregators may be needed to serve as licensing and fulfillment platforms.  And human resources will need to be deployed differently.  More bodies will be needed for account management for broadband and mobile accounts (and fewer for offline retail accounts).  Ditto for marketing and other functions.  The major labels have a leg up on the indies, which don't have the resources to invest ahead of revenue, but lots of work is still needed. 

There are lots of hurdles to get to the level predicted in the study.  But the consumer will continue to demand music in digital form, and the industry will follow.