I have simulated the heat exchange in shell and tube heat exchangers we design and heat exchangers that we buy from outside suppliers. You are correct in your estimation of the difficulty in using Flow or any other CFD to analyze the exchange of heat in a heat exchanger with for example 100 tubes with two, four or more pass configurations. And consider that the heat exchange tubes can also be enhanced on the exterior with "fins" to increase surface area and on the interior with spiral fluting to decrease laminar flow, the question becomes how manyweeks will be required to run the analysis instead of hours.
What I have done is to perform the analysis on just one tube which may also be modeled to reflect 2/4/or more passes. For prime surface tubes this analysis usually takes about 30 minutes. However, if I use finned tubes the analysis will take about 40 hours. At this point I decided that adding spiral fluting, or turbulators, to the interior of the tubes might crash my system before results were even obtained.
Flow can be utilized for heat exchange if your model is very highly scaled. However, you can get fluid flow (no heat analysis) analysis in a reasonable amount of time if the flow is inside the tubes and the inside of the tubes are not enhanced.