The important issue is to distinguish between the observation of
a specific flow (e.g. 22 kg input of steel) and the abstract
flow-object ("steel"). This distinction was not made clear in e.g.
ecospold schema where the exchange = flow and the abstract
flow-object is simply called "intermediateExchangeId" or
"elementaryExchangeId" referring to a taxonomy of "master
data".
I do not think it makes much difference if we call it flow-object
or flow-item, so if there is a preference for the latter that
would be OK with me. Exchange has the problem that it is easily
confused with a market operation (e.g. currency exchange). So I
think "flow" is still the best option, as this is either an input
or an output.
