Note: Mentor Mode is kicking in..
OK, the first and obvious problem in making your robot in CAD and then building it on a FIRST team is time.
Time is your enemy.
In the real world of engineering, a product goes through many revisions and changes throughout it's design stage. (Not to mention the time to design the actual product) On our FIRST team, design work has gotten faster every year and this past year (237's 6th year) the design was chosen by the end of Day 3 or 4 from Kickoff (Maybe sooner)
While I was working as a Draftsman, I was working on one project (an assembly from premade parts and some custom ones and half the size of our robot) for about a year.
This "simple" project had many revisions, and engineering changes.
While in the world of FIRST you don't need to submit the robot to electrical and mechanical standardized testing, you still need some time to wire it up and also drive it.
6 weeks in the long run, is not that much time for design and CAD time..
When we make the robot in CAD first and then build it, we usually just do a simple layout of the robot, or maybe just the chassis for all the drive-train mounting. We like to have a rolling chassis by the Friday after Kickoff (only 5 days of serious designing and building)
If you do the whole robot in CAD you just run out of fabrication time.
*There are some exceptions to this rule obviously. If you have a team of about 5 CAD draftspersons Mentors, you can crank out prints in no time.
Mostly on FIRST teams though, it is a build and learn process. While I could crank out CAD prints when working with the team, it is more important in the long run to teach as you are building.
In the end, I usually teach CAD as we are making the robot and although the process is slow, the actual robot is more important to finish than a 3D model, or a 2d print of one.
(Unless you plan on submitting a Autodesk award entry - A whole other topic)