Foldit team selects diverse solutions from each CASP competing groups and send them to the group administrator, for him to select 5 of them and send it to CASP for group competition. Example: energy, contributors and starting models selected after a CASP puzzle has expired in Foldit)

From the CASP point of view, it seems that the chances of a competing group are higher if we can propose very diverse solutions (for example 1 solution from starting structure 1, another from starting structure 2 etc).

Due to limited resource, we (personally and as a group) always have to choose, when playing the latest days, between:
1-maximizing the group's current top solution (to get the millipoints that will reward us on Foldit ranking);
2-maximizing second best scoring alternative model: more diversity for CASP on the cost of Foldit ranking
3-or even maximizing the lowest scoring model (ending with a diversity of medium scoring solutions both for Foldit and CASP)

Do you think the strategies 2 and/or 3 make sense from (group's) CASP point of view?
Or is it however important (for the group) to get the latest millipoints for a top solution?

From my group's Foldit ranking point of view (and I suppose for Folding team point of view), i would concentrate my latest days on our top solution.
From the group's CASP race point of view, I would concentrate on our second best and following alternative model.