Yup, you're right, and it only adds to my point. The lower rotational inertia (equivalent to mass for rotational systems) of an aluminum flywheel/flexplate will "free up" torque at higher RPMs because there is less torque "robbed" from the engine. This could be seen as an increase in raw measured torque if the engine were strapped to an engine dyno. So you can't have two identical engines with identical torque curves where one engine has a different (lower) reciprocating weight. The one with lower reciprocating weight will indeed produce a tad more torque over time and rev just a tad faster. Now will it be measureable? That would depend on how much load is behind the engine--case in point, an engine in neutral will rev faster than when engaged in gear.