On one hand, there is no difference. People say this to hackers all the time. On Drupal.org this happens so much that there's an official process: There's a specific category in the ticket tracker called "Feature Request" which is applied to tickets in which end users beg the open-source developers to spend their free time developing one feature or another.
Some of those requests succeed. But many of them are ignored, because the number of feature requests and their complexity is orders of magnitude greater than the number of available hackers and their talent.
Of course, if code is built in five minutes for free by a twelve-year-old, it often doesn't work. (To get it to work, you have to pay the twelve-year-old $100 per hour to fix the bugs.) But if a corporate logo is a paint-by-numbers collection of the latest trendy visual cliches, thrown together in five minutes for free by (e.g.) the kind of amateur Photoshop jockey who enters the Fark Photoshop contests... most business owners will never know the difference.
The designers on the no-spec site (e.g. http://www.no-spec.com/articles/why-speculation-hurts/ ) sound like tailors, trying to convince us that mass-market clothing made in sweatshops and sold off the rack is a terrible product when compared to bespoke-tailored suits. And they're right, of course. But from the perspective of an owner of the local pizza parlor who needs a quick logo, it may not make much difference.
I will say, though, that no-spec.com offers some of the most effectively-designed FUD I've ever seen. You should definitely shop there for all your scary-signage needs.
http://code.google.com/android/adc.html
We're still fine, and I'm not mad about it either. It's not like companies like Google and Netflix are forcing people to code at gun point. There is choice involved. If contests cause my work situation to change for the worse, then I adapt (or get lazy and become obsolete...)
In a logo design competition, the elements cannot (easily) be resold or repurposed, since they are specific to the specs of a particular client.
Also, winning one the Android or Netfliz prizes confers lots of value in terms of publicity, which can't be said of 10,000 small logo competitions.
One other note: Rentacoder, for instance, has specific rules prohibiting posting design comps or source code to prevent uncompensated work.
the price of entry isn't as much comparatively and again it's by choice
besides its not like u can't take off the company's name and just include the design in your portfolio. it still has some value
"Make a web application to my specs, please. Do it a few different ways so I have options. If I don't like it, I won't use it, but I won't pay for it, either."
Some industries do follow this model (e.g. architecture, design, custom enterprise SW), but the business case for proposals/POCs/loss-leaders can only be justified when:
(# of proposals generated) x (cost of work to produce proposals) < (value of awarded contracts) - (cost to fulfill contracts)
However, with contest sites/spec work, that equation only makes sense if you value your time (costs) at close to zero, since the ratio of proposals generated to awarded contracts is so lopsided.