Mentor-ship is the biggest part of being elite. If you have mentors who know how its done and how to teach kids how to effectively and efficiently design and prototype then you can achieve almost anything.
However, if you can't control mentor-ship (because you just can't) then be the student that just takes charge of it all. Common things that my team always takes into account in competitive though is:
-First(not FIRST*), you are not trying to just win. You are trying to be professional. Think of your team as a company designing a product for high paying customers(also known as sponsors). It's supposed to work effectively while still looking like something made by a high paid professional.
-Never design a robot before deciding what it has to do.
When the game is released have all students read every rule and come up with strategies. 2010 was easy for finding strategies, front zone, middle zone, far zone, etc. By doing this you can then find what you think is the best way to play the game instead of, "Oh, but I already designed this robot..."
-Never design a robot (during season) that you don't know how to build.
This year 973 went with an arm even though we knew that an elevator would probably be more effective. Why? Because we have zero experience with elevators and to design an effective one from scratch during season is way too much pressure. This allowed us to focus on other things such as the minibot, manipulator, and drive-train.
-Have a set of design patterns that you follow with all parts you make. Some things that we always follow when designing any system include: Everything that can be deployed can be undeployed (no way to get around the minibot), hole pattern on all tubes(usually on 1" face), design all parts from the machinists point of view(you are more likely to keep sponsors when they can just make the part without having to call or do any math), use the same size bolt everywhere(its easier to repair the robot when you don't have to switch tools), etc...
-Build two robots. Not just for practicing driving, but to better understand how the comp bot will go together and what(if any) changes need to be made.
-Don't be afraid to use someone else's idea. Unless you claim it as your own, many elite teams won't care if you use their idea, even if it looks exactly the same(well, at least my team wouldn't care...).
-SCOUT. Scouting is as important if not more important than your actual robot performance. Even if your 90% sure you won't be in the top 8, good scouting helps you win.
Hope to see you at champs next year.
