Then the questions to really ask are:
If they ARE having trouble accessing Skype, as they and several other "super-secret" agencies that never talk about what they are doing claim, why are they advertising that fact? To draw subversive elements into Skype so that a year down the line they can go... well, actually, we could monitor it - thanks for the info?
If Skype is *really* that good (I doubt it, it's probably nothing more than a nice AES with some custom tweaks or other encryption layered over the top), how do they expect anybody to find a solution?
If the encryption *is* that damn good, then you have a copy of the encrypter/decrypter code inside every Skype executable. You might not be able to get people's private keys but the protocol is basically published inside every executable, which are available in a variety of platforms and languages. Ever heard of a disassembler, guys?
I suspect it's more the P2P element that flummoxes them because there is nowhere sensible that they can intercept random calls. The fact is that if they suspect anyone, they can easily target them specifically (keyboard-capture, screen-capture, virus infection, social engineering, etc.) and it's a lot easier than trying to monitor the whole world. But for some reason, governments always want complete control of such things - probably because that is the best way to subvert and hinder your citizens (ala China).
The problem they have is that Skype is pretty much a distributed system so they can't do their mass analysis and come up with "X is a terrorist" just by watching random traffic from everybody in the world (that is, of course, if they ever could).
It seems to me that this and the GCHQ affair are aimed at a political change, not an actual valid call for research. They want the power to monitor everyone for everything so they can pick up on the same trends that they used to be able to. If they say they "can't" do it, maybe the government will ban it, or enforce an alternative, or force backdoors, etc. With P2P, point-to-point encryption with well-known and tested algorithms, it is virtually impossible to break even a simple message and they have to go back to old-fashioned policing - work out who the terrorists are by watching known terrorists and following what they do.
Personally, if this is the case and they have no way to watch it - good. It provides a bit of anonymity and security to the network again. I don't see any reason that any modern protocol should NOT include encryption nowadays, it's so cheap to implement and can apparently stop even these "big guys" in their tracks even when they know the entire protocol and have access to public keys.
This is what happens when you try to control things that don't want to be controlled - they find a way around it. Each time you control something else, the ways around get smarter and more powerful. And before you know what's happening, the entire IPv4/IPv6 infrastructure is nothing more than a carrier for complete encryption for everything that everyone does online.
People already KNOW that wireless has to be encrypted and that secure website are required if they pay by credit card. Soon, the whole internet will be nothing but a huge P2P VPN. And then where will the overbearing legislation get you? You won't be able to stamp on Joe Bloggs because he has encryption technology you can't beat (ala the PGP etc. problems of the early days) and to use ANYTHING online (which is what every government is pushing towards), you'll have to be part of the same network that the government hates because it can't break it, censor it or monitor it.
The Internet has leapfrogged these sorts of organisations and "Big Brother" ideas and they can't handle it. Unfortunately, it was caused (at least in part) by their overbearing manner in the first place - chasing kiddies for downloading songs, etc. They can't break it? Good. Let 'em whinge. While we're at it, let's see how much more terrorism happens without them.