Hmmm... the dodo option would minimize the amount of chimeric additions needed, I would think. We could thus test out the approach more cautiously. And bringing back perhaps the most famous symbol of human-caused extinction, if marketed correctly, could garner support from unusual quarters.
That said, it sounds like Wu is eager to really stretch his technique to the limit (he was disappointed to not go for dinosaurs right away), so the dodo might be a bit too cautious of a step on that front. As long as we don't go for the massive species, they're unlikely to consider humans potential meals; 1-1.5 meters should be reasonably safe on that front (although keep small children away).
Keeping to birds in preparation for meeting in the middle is probably better than going for the curveball, at least for now. Terror birds to start with, then reach back for some other unique-looking birds, finishing with a species or two of archeopteryx.
On the head zookeeper... hmmm. Not honestly in favor of the "hands on" option, as ultimately the species we'll be bringing back are likely to react in ways outside modern experience. Automation is tempting, as is the ability to make nice with one of the local groups.
On systems... Stuart appeals. While the potential of closer US governmental eyeballs could be problematic, if we're careful about how we use our techniques it'll probably not be a major concern. And if there does end up being an "information pipeline" established, that could actually provide reassurance we aren't trying to hide from the US government. The "rock-solid" description also reassures me of competence. The BioSyn option could be useful on the espionage front, but not sure if it's worth the hassle. The eccentric genius is a gamble in my opinion; if he could be induced to buy in to the vision, he'd be spectacular - but otherwise, he'd end up a distraction.