There are times when reasoning by analogy isn't optimal. Elon Musk famously preferred going through "first principles" and I find this likely leads to a better more innovative end product.

Take payments (my industry) for example. There are countless startups that are working on something that enhances or optimizes whatever pre-existing solution we have today. That's thinking by analogy. It's faster and easier but you're constrained by the limits of whatever you're comparing to.

Then there's thinking from first principles. From scratch. Take what Dwolla is trying to build for example. "If we were to build the Visa or Mastercard network today, what would it look like?"

A lot of the new innovative systems in payments, like Bitcoins, started by thinking of first principles rather than analogy.