This also goes for many things in general, not just gamedev. I used to be a teaching assistant at the University that I was studying at, and this was the main thing people seemed to get wrong in their projects. Instead of going for the basics and building from there, they just went for all the fancy cool features, or the most optimal algorithm. Then, when the deadline inevitably came around, they would have basically nothing working correctly. Sometimes I even warned them, and yet it still went wrong.
Right now I’m basically playing Beyond All Reason almost every evening. It’s a game in the Total Annihilation “tree” of games. A massive scale RTS. I previously played Supreme Commander and Planetary Annihilation, both of which are also inspired by Total Annihilation, but I have to say that BAR is really better than both of them. I almost can’t believe it’s an open source game. It’s still in alpha, but it’s been way more stable than most AAA games I’ve been playing recently.