I will argue that some projects are definitely overly architectured simply for the sake of gaining experience in doing so. If everyone waited until they actually had Google scale problems to learn how to create scalable architecture, then only a small subset of people would have that skill. It is the initiative of doing so and ability to point at previous history of successfully completing such projects that give these people the opportunity to work on problems that actually require said complex solutions.
The one thing that I've learned by being surrounded by incredibly talented people is that you can't compare yourself to others in a singularly dimensional way. If you do, the odds are against you that you are the best in the world (1 in billions, in fact). This is the quickest path to defeat and nihilism.
What I've chosen to do instead is compare myself across all dimensions that I have interest and passion for. The beauty of this is that the combinatorics of genetics normalized for individual circumstance leads to only one meaningful comparison, yourself. Internalizing this just leads to a more fun game.
It's probably more accurate to say that it's a quadrant with the axes being 'comparison' and 'gratitude'. Being high on gratitude and high on comparison accounts for your quadrant of possibilities. It seems like the moral is to either be lower in comparison, or higher in gratitude. Or both :)
