> Trouble is, the whole thing has to be safety tested for a
> manned environment which adds so much delay and cost it
> makes the whole thing pointless, or nearly so.

...there are special risks and costs associated with doing a mission either way. Some risks and costs associated with not using STS/ISS (some of the following assumes they want their rats back):

- must supply own power.
- possibly must supply telemetry that is otherwise unnecessary.
- must survive in space.
- must survive much rougher handling (launch/re-entry).
- must purchase or build reentry propulsion/shielding.
- must fund retrieval of the experiment.
- must accept risks of non-man-rated launch and retrieval systems.
- must accept risks of their not being a handy human to supply a minor. (but crucial) tweak to the experiment ("Give it a 0 G kick and see if it starts!", eg.).

Did I forget anything?