Xorak posted this very useful comment in another thread. It's worth it, reading about cell count. Helped me a lot. There are a lot of other helpful post on this subject aswell.
Quote Posted by Xorak
I've been spending the last few weeks doing this for my own mission. I've now gone from 2,500 brushes and 30,000 cells, to nearing 5,400 brushes and 32,600 cells. I’ve more than doubled the brush count and might even reach 6,000.
What I've done is convert most of the cylinders to objects, mainly the ones the player won't interact with directly, like say the molding around a high ceiling or something. With a nicely applied texture, there's no way a player would know that's an object and not a brush. With enough time and care, even any square/rectangle brushes you use for detailing can potentially be converted to objects. If you're crazy enough, you could build the entire mission without using a single solid-brush.
Another thing I've done is remove wherever possible the places I split a room into two or more stacked air-brushes (where each brush had a different texture applied to the wall.) That wastes so many cells. Instead, I’ll make the room out of a single air-brush from top to bottom, and apply thin solid-brushes, with the various different textures, to the walls like band-aids. That method creates no additional cells. It drives up the brush count, but so be it, you're allowed thousands more than you'll ever be able to use.
Probably the most tiresome thing is fiddling around with the Time of your brushes. You’d be surprised how many cells you can save doing this. In one case, I 'To Ended' a single square air-brush and it saved 100 cells! You generally want the most complex brushes to have later times, so the peaks of your towers and such should generally be the last in Time order. But this is also where it gets tricky, because I've found it doesn't always work as simply as that. I'll sometimes 'To End' a brush that is certain to save cells and it instead creates more. So right now I’ll ‘To End’ a brush, optimize, and if it recovers cells then I save and move on to the next room, but if it doesn't then I reload and try the next room. If you absolutely must push Dromed as far as you can, this is the sort of stuff you have to be prepared to do, I think.
Also, in your mission, I noticed you have a lot of windows inset into the walls. With a nice enough texture you can create the illusion of the window being inset while not having it be so. You’d save hundreds of cells there if you made the windows flat to the wall, though you might lose a bit of the atmosphere. So it’s a trade off depending on how desperate you are.
It certainly helps also to carve everything out of air-brushes to begin with, rather than having huge air-brushes in which you then add the solid buildings into. This alone saves thousands of cells over the long term. This doesn't really matter too much if you're only building a medium sized mission to begin with, but if you want huge, then you'll probably have to build in this way.