My counter argument is: "Good Ideas Grow On Trees"
Working with a startup myself, I can clearly see why Paul Graham nailed it down to "growth" and not "idea".
The "idea" part of an startup is the easy one. I scientifically researched the topic 'innovation capability' and 'innovativeness' of organizations and the crucial part is not to come up with an idea. As said "good ideas grow on trees". The hard work is the execution of an startup idea to get "growth".
The main difference between startups is not their idea, but their growth. Many people can come up with a new startup idea in days, maybe even with a new idea within minutes. But it can take years to bring an idea to success.
Here are examples of good ideas that are growing on trees: http://www.quora.com/What-is-the-best-way-for-a-first-time-e...
There are endless startup ideas. Even the idea newness, how radical new a startup idea is, is not the main aspect of an startup. Again the main difference between startups is if they can execute the idea and get growth. For example with our startup we were the first to hit a large, untapped or underserved market. But the lack of execution and growth was what killed it. There are always other people who sooner or later have a similar idea. The main difference again is which startup executes the idea the best, the fastest and gets the most growth.