There is an interesting 3-page paper from Donald Geman, a math professor at John Hopkins University, on "Ten Reasons Why Conference Papers Should Be Abolished". He is basically arguing that there are not enough quality results to justify the number of possible conferences where research can be published. The whole publication cycle is being demeaned with poorer results receiving poorer reviews, for which there are fewer people with time to read the final outcomes. He even suggests to limit your lifetime publications to 25 or so in total.
There is certainly an arms race of sorts at work here. In The Crypto Year in Review from Bart Preneel (the year in question being 2009), Preneel observed that cryptographic research is alive and well, with over 600 papers published on the pre-print server of the IACR, amounting to over 10,000 pages of written research. The current count for 2010 is 450. And that is just cryptography (high math content), and the number of more general security papers will be much larger.