The best things I've learned from 37signals are

  - You can build cool things that people need and will pay for.
  - You can keep it simple and effective.
  - You can do it in whatever style suits you best.

Copying's got nothing to do with it.

"Bad artists copy. Great artists steal."

When you learn, you steal. ;)

It is an art form in itself to copy what's working rather than what's incidential to the success of a firm.

Learn from your peers in industry, sure, but make sure you take the lessons to heart and make them your own rather than just engaging in blind cargo cult imitation.

When you know nothing copy. Artist have been copying each other for centuries. The Roman's copied the Greeks, the French copied the Romans, etc. I think when you start out you have no choice but to copy. As you get better you can start to do things differently adding your own to it. Eventually you are the one innovating.

Remember copying alone never pays off, but it's a great way to learn quickly when you know nothing. Microsoft was the best at this back in the day. First version was always a copy of what existed out there. After that they got feedback and iterated on that design surpassing the incumbents. Eating your own dog food helps in this process. They did this over and over with lots of success.

As noted in these posts, copying you copy success and failure. It's your job to sort which is which for your customers/community. You have to know the point when copying won't help you, and innovation should take over.

"2. Be your own customer. Try (if you can) to eat your own cooking. A product works out much better when you use it yourself. Solve your own problems. Fix the things that annoy you the most. Beyond just 37signals, there are lots of examples where people built software that succeeded in part because they use it themselves. GMail comes to mind."

This is kind of ancient wisdom in the software industry (even Microsoft does this), but as you say, it may be worth repeating.

I have a biased opinion (I wrote the article), but I'm not sure that advice necessarily needs to be "fresh" in order to be useful/good.

I learn "new" things all the time that it seems others have known for a long time.

I don't agree with "5. Charge early, charge often." as being one of the things a startup SHOULD copy from 37Signals. Sure, it has many benefits, but it's just not feasible for websites that heavily rely on 'network effects' or user capital to provide value.

The chicken and egg problem is hard to overcome already, and charging for access early obviously doesn't make sense for these types of websites.