Back in the 20th century, every looked at the incredible advances in travel technology from 1850-1950 and assumed the growth curve would just keep on going. But it didn't; it peaked out somewhere in the 1970s. Since then it's gotten cheaper, but not faster.
I think the Singularity is the exact same sort of projection. Progress in computer technology was incredibly fast from 1970 to 2000. My phone has a processor thousands of times faster than my first computer, back in 1982. And if progress continued at that rate, the Singularity probably would be inevitable. But it hasn't. My desktop computer today is not significantly more powerful than the machine I had in 2007. My laptop is much better than my laptop then, but it's only on par with my 2007 desktop. My phone is insanely better than my 2007 phone. The current trend is for computing power to get smaller and cheaper, rather than getting more computing power. That's great, but I don't see it getting us to the Singularity anytime soon.
I think it's a mistake to compare specific devices from different eras, like a 2007 desktop and a 2012 desktop. In order to find total computing power, you need to add your phone, tablet, desktop, laptop, and cloud services together. When you do that, you see that we all have far more computing power than we did in the past, but we've chosen to spread that power over a variety of different devices.
Your desktop may not be getting much faster, but your TV got smarter, baby toys have more computing power than your first PC, a new car has more smartphone features than your phone from 2007, and everything else seems to be getting more computing power whether it needs it or not.
The Big Science stuff is cute and all, but don't for a second pretend that technology in common use isn't mindblowingly awesome.
http://en.wikipedia.org/wiki/Motorola_StarTAC
Everything discovered in CERN today could have been discovered 20 years ago.
http://en.wikipedia.org/wiki/Superconducting_Super_Collider
Yes, because of semiconductor advances (Moore's Law) computers and communications advance at a blinding pace. However, we would all benefit if we could push other areas too.
It's not about how far we've come, it's how far we have yet to go.
Anyway, the one good thing that could come out of the Singularity discussion is maybe we can have a concerted effort to make it happen. A lot more research money into big science, for example. Hypersonic flight, maglev trains, the Texas supercollider, space exploration, medical research, clean energy, etc. Imagine if the the first flight happen 100 years before the Wright Brothers, or the telephone had been invented even 50 years earlier. We might not have a Singularity by 2045, but we can actively accelerate our quest for knowledge.