Actually, you can, which was the point of his essay: focus on validating your vision, rather than being reactive to someone else's actions.
They were too focused on being reactive to a competitor, rather than really understanding their users. Because if you just copy whatever your competitor is doing, you're getting the understanding of your users second-hand, and you might end up copying things that aren't important--you might end up cargo culting.
He's not saying, "Oh, if we had only just done photo sharing, we would have knocked it out of the park!" He's saying that while checkins was a way to validate that users want to play a game that lets them see the world through their friends' eyes, they should have been open to other solutions to validate against that original thesis (like photo sharing), instead of being caught up in being competitive overriding the validation and learning about their users.
The harder part is really looking at your users, how they use the product, what they want - but not what they say they want. With my product, we have people requesting features of the 'big dog' - only because that's what they've seen. We're not about to blindly implement what a competitor does just because they did it - it's assuming they did the research into that feature, and we don't have to.
Anyway, essentially - it's really hard to not copy a competitor when that's all your users are vocally asking for is to do just that.
They were too focused on being reactive to a competitor, rather than really understanding their users. Because if you just copy whatever your competitor is doing, you're getting the understanding of your users second-hand, and you might end up copying things that aren't important--you might end up cargo culting.
He's not saying, "Oh, if we had only just done photo sharing, we would have knocked it out of the park!" He's saying that while checkins was a way to validate that users want to play a game that lets them see the world through their friends' eyes, they should have been open to other solutions to validate against that original thesis (like photo sharing), instead of being caught up in being competitive overriding the validation and learning about their users.