Eh, I graduated in 2008. ENCODE was only a pilot yet. We didn't know about lincRNAs. My estimate was based on multi-species sequence alignments so it is pretty conservative. At the time it was not, however; it was doubling the amount of functional sequence. To be honest i am not sure what the current best estimates are based on (i switched fields). 8% seems absurdly high to me, but there is a ton more evidence now than i had.