You've read bits of my post on the matter, so you already know my answer is "the optimal bidding strategy is unclear." I hope that this ends up meaning that on average, all strategies are equally good.
I suspect that all strategies will basically average out to the same return. Particularly on days 6-365, waiting until the end of the day gives you the advantage of knowing when not to contribute, but you still don't know for sure. The new bidder safeguards that I've heard about for end-of-day bids should help end-of-day bids not to be too bad.
On the other hand, even if the auction is competitive after the first 5 days, the crowdsale price can't rise above the market price on average - so there's no downside (again, on average) to contributing at the start of the day.
Arbitrage will wipe out any systematic advantage that the secondary markets offer. You might get lower-variance pricing on the secondary markets, but mean pricing should be the same.
Of course, there very well could be some cognitive biases at play which will make one of the above strategies significantly better than others. In addition, there are other forces at work. EOS employees, for example, will probably favor buying in the crowdsale over buying on secondary markets - which could push crowdsale prices above secondary market prices.