Hi,
The crux of these issues is discovering or making a workflow pipeline. After that, you can go to the several online communities involved with the software creation like your pipeline and get more information.
Of course, once you settle on the basics, you may come here for specific questions and get precise answers.
I've also heard that simple color palletes (maybe a little more saturated too?) help achieve a clean look.
What is a "clean look" can be attractive to some people and boring to others. It's a matter of preference. A little market research into what exactly appeals to your target segment of the market is crucial or you will be mostly guessing for quite a while. This is why beta testing and demo back feed are so important.
Of course, some things just beg for a "clean look" such as the paint finish on a sports car, for example, or the face of a clock or instrument gauge.
Also, what about keeping a room a singlular hull of geometry? is this really neccessary?
That depends on different factors such as performance demands, time (and therefore budget) available to work on it, UV mapping ability and speed of the artist or brush skill in your case, and the variance of the environment such as terrain elevation variety which can force you to more customize the shape of things.
Only experience can resolve some of these things in you.
Any other tips would help a great deal. I am between intermidiate and noob with Maya and Zbrush when it comes to flat out modelling stuff.
(i know how to rig and animate).
Though I don't use Maya and Zbrush, they are great. Workflow pipeline demand has taken me to other software, but the general principles are the same.
1) Try to duplicate surfaces and UV maps as much as possible to help performance issues in a game.
2) Keep polygonal count of every model and part as small as possible and depend more on texture appearance rather than geometry.
3) Use 2D surfaces instead of polyhedrons as much as possible.
4) Keep ray tracing and path finding to the minimum that you can.
5) The fewer the files in the model folder, then the better, such as material files, animation files, ray tracing files, color files, brush files, and so forth. Keep the size of these files as small as you artistically can, too.
6) Keep a permanent storage of your work in progress (WIP) folders for future proof of your work and taking components for future projects, both 2D and 3D.
7) Make a WIP stage copy of your work every 10-20 minutes incase you make a huge blunder or your system crashes, as well as for proof of creation records.
8) Do not reinvent the wheel, so to speak, but reuse and modify already existing art work as much as possible within your license rights.
9) UV mapping, 2D image manipulation (such as GIMP), and brushes should all be used in 3D games, depending on artistic preference and performance issues.
10) Learn the positive habit of doing quick "over once" work and then return in the following cycles of art creation to refine each artwork. This can also apply to the game as a whole when you get all your basic "place holder" unfinished assets into the game to test for performance before you go to the next cycle to finish the art assets. This saves much time in the long run due to realizing exactly how large and how many assets is appropriate for performance and appearance in the game. For example, if I made about 20 too many types of buildings that game performance would allow, then I just wasted a few weeks of time. 
LOL If I go and finish each model before I do a game test of all the models together in the game as unfinished place holders, then I am at high risk of having to remake models, as well.
11) The goal is "made right, first time every time". Though not always possible, this is most efficient and the industry standard for artists.
With 3 1/2 years in the game and simulation development industry, I feel that this set of advice is trustworthy. 
