For comparison of flow without, and with, a reservoir, you should be able to simply compare the output from the the rch and rsv output files for the same subbasin. Let's say your reservoir was in subbasin 15. The flow_out of reach 15 should be what is going into the reservoir (flow_in), and this should be the same whether the reservoir is there or not. The flow_out of the reservoir then represents the result of the reservoir in modifying the flow, depending on how you parameterized the reservoir. I've only used the "targeted release" option because it's the only one that seems to give realistic results for my situation, where most lakes and reservoirs allow water to pass through without much control.
For Ponds and Wetlands to have a hydraulic-storage effect in reducing storm peaks, you probably need to set your IFLOD months so that the model thinks it is "flood season." During flood season, SWAT allows the water volume to drop below "emergency" volumes, and so be released gradually (based on the NDTARG parameter) down to the principal or normal volume. Otherwise, during "non-flood" months, SWAT will allow the water volume to remain as large as possible, at the emergency volume. If Ponds and Wetlands are at their emergency volumes, then any added water from stormflow runoff gets spilled immediately, and the result is that there is no amelioration of flood peaks -- just as if the Ponds and Wetlands weren't there. (This happens under humid climates where water supply exceeds evaporative and seepage losses -- it could be different under arid climates.)
Just FYI -- you can easily "turn off" Ponds and Wetlands by setting PND_FR and WET_FR all to zero in the pnd table of the project database. Save the original values so you can update the table back to its original state. This way you can compare the output with Ponds and Wetlands turned off, and on.
-- Jim