Still, I think Kendricke's advice should be qualified a bit. A good designer (of anything, not just games) doesn't just come up with new ideas but doesn't just duplicate good ideas either.
Perfect the good
First, no idea is ever honed to absolute perfection. Ludwig van Beethoven put it well:
"The true artist is not proud. He senses dimly how far he is from the goal, and though others may admire him, he feels sad not to have reached the point where his better genius lights the way like a distant sun."
We will always fall short of the ideal, but that's no excuse to quit reaching for it. In game design, reaching the ideal refers as much to finding which features fit best together as it does finding the best possible iteration of that feature. No feature can reach its full potential if placed into a poor context. A poor context might include many other great features... but the wrong combination of features.
Second, there's rarely only one form that can capture an ideal excellently. Beethoven's "Moonlight Sonata" and "Pretty Good Year" by Tori Amos (or insert whatever song and artist you prefer) both succeed admirably on many levels, particularly to communicate sadness, but they accomplish the same goal in different ways. In game design also, a design may have worked excellently before but there remain other ways to accomplish the goal.
If it ain't broke, don't fix it. But we shouldn't automatically steal any idea that works. Consider using that old idea, but also consider if there are other ways which better fit your overall design or which you, as an individual and unique designer, might be better able to realize.