in order to calculate transit delay values, as you've put it, you must include all the devices end-to-end in the communication path.
end user to the service may transit multiple routers, switches and media types. all of the induced delay/latency of these need to be taken into consideration for your overall picture of transit delay.
also, the different platforms have different capabilities, speeds and processing power. these attributes along with the individual module/interfaces must be included.
i would take all these points end-to-end, determine their individual latency and add them together for the end-to-end trasit delay. do not just compare one chassis to another as this may not cover all the pieces in between along the path end-to-end.