I've stopped bothering with industry efficiency rates. I've found that nothing is going to keep your goods cheap once inflation gets going. The only time I ever invest in efficiency is when I have a large MG manufacturing base I want to sell off for a while to come. Even that, however, eventually falls off.
As for immigration, I've never bothered to do anything with that either (I don't seem to do that much, do I?

). I generally just let as many people into my country as possible and then try to recruit (or conscript, depending on my manpower situation) them into my military.
And for what I research, it's whatever I want to. Usually my first order of business is the range-boosting artillery upgrades (V-LAP, EFRB-BB, etc) since that fits in most with my fighting style. After that, I usually go for the cheaper social programs, then work up to the more expensive ones. If I need certain classes of units, then I research those as well. However, I usually attempt to pump at least $100,000M into my research per year, but on smaller regions that can't support that, I can understand where maintaining research efficiency and tech level work would make sense.
Oh, and don't pay too close attention to what the internal-external numbers are saying in the treasury screen. I've had an internal deficit of 20-30% more than my exports total and still made a profit on the year. As long as you aren't totally scraping bottom, keep raising research funding. If you hit a cap, build another center!