A difficult one. Are you looking at a single endpoint determination (eg is the cell line infected with xx after treament +/- y drug? or are you looking at a variety of different parameters eg a panel of cytokines/chemokines, changes in expression levels?. Are you looking at absolute number quantification (calibrated against a plasmid standard) or relative eg fold increase.? What scale are we talking - 100s, 1000s? samples gathered in real-time/ processed in real time ? or archived? Do you have access to any form of automation? these are all factors that can have a bearing on the technology you use. What is the timescale? Do the samples have to be processed/ quantitated at the same time or is it a longitudinal study? will you batch process?
Doing / processing 20 samples a day over 2 weeks can be done manually. Doing 200 samples in 1 day is a different matter. Costs will be dictated by the technology you use - are you intending to use kits? PCRs in microtitre plates or in individual tubes? One step or 2 step RT-PCR?Multiplexing? generally time = cost and less time = more cost.
My advice is to draw up a spread sheet of your likely experimental sample throughput and then work out how much can be processed by an individual at one time and then go for the appropriate technology. Trisol extraction in my experiance is the Gold standard for RNA extraction, but is not the ideal choice if you are going to process 200 samples in a day.Buying bulk cDNA synthesis kits is fine and cost saving, but are you going to be processing the samples over 6 months using the same open kit? Etc Etc must stop ranting ......