There’s a number of things you can do.
The first is to recognize that even VETERAN teams don’t always understand that focusing on one task can be pretty important (148 in 2008 as my witness). However, good veteran teams recognize that you need to do the entire game, and will either design in sequence (that is, what’s coming up next must work around what’s already there) or as a unit (everything at once). (And there were quite a few rookies that did both gear and climb.)
Second, building on that, suggest a “sequence build”. That is, let’s just pick on Steamworks: Build the gear mechanism first. Then, once that’s well underway, start spinning off members to work on the climber.
Third, you need to start increasing the team’s knowledge base. Find your local offseason workshops, and drag as many team members as you can with you. You won’t regret it, or maybe you’ll regret the lack of available time to attend all of them… Can’t attend the workshops? Find out who is presenting, and ask if they can give a lecture or two at your meetings. No local workshops? Post back and I suspect you’ll be slammed with videos of workshops…
Fourth, apply that knowledge from the workshops. Build a summer (fall) robot next year and attend your local offseason event with it. Build a demo robot. Add on that fuel shooter to your current robot. Test stuff out.
7-8 dedicated members isn’t a problem. It’s an opportunity. Let’s say that for next year, you plan to do two mechanisms. You take those dedicated members, and tell them to grab minions (non-dedicated members) and get busy working with them, and you tell the minions to learn what they can on the fly. Ideally, you do this during the offseason, but the general idea is to train those less-dedicated members and get them more involved so they realize how much fun this is and become more dedicated. Then the following year, repeat the process, but with another mechanism (let’s just call it an alternate prototype for doing one of the game tasks).