Most people now know that ENCODE did not disprove junk DNA (with the possible exception of creationists and a few kooks).
Sean has written a wonderful article for Current Biology where he explains in simple terms why there is abundant evidence for junk (i.e. nonfunctional) DNA [The C-value paradox, junk DNA and ENCODE] [preprint].
Here's a quotation from the article to pique your interest.
Recently, the ENCODE project has concluded that 80% of the human genome is reproducibly transcribed, bound to proteins, or has its chromatin specifically modified. In widespread publicity around the project, some ENCODE leaders claimed that this biochemical activity disproves junk DNA. If there is an alternative hypothesis, it must provide an alternative explanation for the data: for the C-value paradox, for mutational load, and for how a large fraction of eukaryotic genomes is composed of neutrally drifting transposon-derived sequence. ENCODE hasn’t done this, and most of ENCODE’s data don’t bear directly on the question. Transposon‑derived sequence is generally expected to be biochemically active by ENCODE’s definitions — lots of transposon sequences are inserted into transcribed genic regions, mobile transposons are transcribed and regulated, and genomic suppression of transposon activity requires DNA‑binding and chromatin modification.
The question that the ‘junk DNA’ concept addresses is not whether these sequences are biochemically ‘active’, but whether they’re there primarily because they’re useful for the organism.
1. More importantly, he's an alumnus of Talk.origins.