ok, well in Sandbox I only play Gigantic abundant maps and I too have run into the money issue. I have played at Suicidal -though no victories yet

so I know what you are talking about when you sit there and watch the AI launch massive amounts of colonizers faster than you can blink.
My play style: I use a MV-capable custom race that has IIRC, 65%morale (allows for a starting 57% tax at 100% morale, I can usually hold it at 49% for 100% morale, which is needed for the pop. boosts) 40% Econ, 5% Pop. Growth, and 20% Social Production, Super Hive ability (the one week build time is a gift from above). I go with Universalists (for the small econ boost and pop. growth boost, but mostly cuz I can still get my Luck bonus, needed during fights)
While the HW can be used as a major econ center, its more important in the higher level games to be getting those ships out there, whether it be colonizers, constructors, or military, I build up my infrastructure as much and as fast as possible on my HW. Be SMART about rush buying, try to avoid rush buying on an undeveloped tile, and also, if you can wait ONE week before Rush buying, it drops the cost a good amount. Finally, be cognizant of the leases. They are (well the Mitrosoft anyhow) very useful but that 20bc/wk cost adds up QUICK so if you make a purchase make sure you can make that kind of return on your investment.
I don't tech trade. I never get the kind of money I think I should and I really loathe giving anything away that diminishes the distance between the AI and myself in tech. IF I do have tech trading on, its OK to trade with minors, actually I usually try to use tech and whatever else I have to get the Treaties from them, and I also like freighter trade with Minors. Basically, look at them as free money, maybe not a lot of money, but especially early, enough to keep you afloat.
DON'T back fill. I know it sounds tempting to get those borders set up and one would think, ok well this one is 4 wks away and this only 1, I can get to it at anytime. The AI will already be settling in on your door steps if you do that. Your better off taking the worlds close to you and getting them to make money than taking something furthur away.
I make all Class 10 worlds and less automatically Econ-based worlds (and worlds that are higher than 20, paradise worlds like that, sound so tempting for lab or factory centers but take FOREVER to build up and end up costing a fortune)
Ok, so the Drengin have Toxic worlds and the Iconians have their Adapter and Yor have Barren. Big deal. I will usually put off Extreme Colonization unless I notice that I am majorly falling behind in the Planet race and thats the only way I can get ahead. Extreme Colonization techs are just too expensive at the beginning of the game to really be worth it (since you are looking at three techs just to make a world viable)
IF you do go for the Extreme Colonization techs only go for the BASIC level and leave the advanced for later, use those worlds strictly as econ/people boosters until you can really put the time into research. By the time I would have done all my Extreme colony techs I could have had Cultural Conquest and just influenced worlds into my control

DON'T worry about debt. I just played, beat, and got a conquest victory on a Tough map that I was in debt for the ENTIRETY of 2227. You can drop to -500BC before your manufacturing shuts down, at which point, at least if you are using All Factory, you will have a surplus, bringing you back out of the red. Obviously, the Stop/Go isn't a good thing but usually within 10 turns of originally going in debt on a Giga Map, I am successfully showing a surplus. I tend to run about +4-8K with 100 worlds, but I also don't maximize their economic potential like I could (9Bil worlds, 49% tax)
Finally, somewhere above me, someone talked about troop transports and using those. It IS possible to have Planetary Invasion and be ready to take on the AI worlds as early as July/August of 2226 (in DA), just have to be smart on what you research. I will say going the troop rush route works MUCH better with an All Labs approach.
Just some various thoughts.