I've been working on these tickets. E.g. Screen-shot: http://trac.openmicroscopy.org.uk/ome/a ... .14.53.png
One question: I've implemented the "load defaults" behaviour to only load defaults for parameters that have a default value AND the parameter value itself is empty. This is to ensure you don't accidentally overwrite your data when loading defaults. If you want to overwrite, you can "clear experimental values" first.
Does this behaviour sound OK?
The only potential probem is that you may have some parameters which have experimental values but no defaults and others with experimental values AND defaults and you want to overwrite with defaults. You'd have to clear values from the whole experiment to 'overwrite', and this would also clear the experimental values from the paramters with no defaults.
The other option is to ALWAYS load defaults if they exist, even overwriting existing values. OR allow the user to choose, which means asking every time, or giving the user 2 toolbar buttons - one for each behavior. Does that make sense?
How does this affect your use-case?