I work things by priority. I point at a project and say "I want this", then find the best researcher (lab count vs bonus for highest possible RP) and put him on it with as many labs as I can allocate.
Repeat for next wanted tech if I have enough leftover labs.
Then every other researcher unassigned, I find jobs for them to do and give them 1 lab each. The highest effective bonus (for the x4 specialization) goes to Research Rate, 2nd best to Construction rate, 3rd best to Economy; unless they are one of the focuses. After that, whatever I feel like doing.
Often, I maintain a task queue outside the game for the techs I want.
Rule of thumb:
Engines tech is considered baseline
Research Rate should be baseline+1, when increasing baseline, Research Rate goes to baseline+2 before engines
ECM/ECCM is baseline+1, behind engines
Construction rate and Ordnance keeps pace with baseline, engines come first
Weapons, EM sensors are baseline-1
Active sensors are baseline-2
At game start, economy is at 0, Economy is priority until deficit is corrected. If deficit occurs again (or projected to occur), Economy overrides everything else.
I use the same rule for construction queues. Enough to complete in one increment or 100% whichever is lower for first task, spilling over as needed.
Construction factories keep pace with population, built in batches of 50-100 as required. Goal is to have 100% employment with no penalty from overwork.
Ordnance factories are built to keep pace with ship production. Goal in peacetime to have ~2 complete fleet reloads of latest + TL-1 design in service (so one reload on planet, one in ships). This goes up as reduced size launchers come online to around 5 reloads.
Automated mines are produced as and when I require, this take priority over everything else.
I often find a planet or moon with no useful deposits and make it my research planet. RLab production keeps pace with population on that planet.
All leftover BP is component production or shipyard production (if all are full).