For Google, CTR is king (everyone wins: no clicks = no one gets paid) and because your account history clock is always ticking, you need to get highest CTR as quickly as possible.
So the strategy should be to bid aggressively right from the start to get a high CTR and then start to bid down to optimise for lower prices and probably better conversion at lower positions.
Optimally you should be tracking conversions and able to take advantage of Conversion Optimizer (30 conversions per campaign per 30 day period or less) – it works really well in all the testing I have done and read about, but the rules of the game are very different.
Yes, you can “lurk” with low bids and low positions just to gather impression volume data (I’ve done this) before you commit to higher bids and positions for selected keywords.
However, you need good Quality Score to get away with this or your ad won’t show and gather the Impression data you need, and over time your QS is likely to be adversely affected by a low CTR.
Certainly for the Content network, Google have published case history data which dramatically evidences the need for initial bids to be artificially high to get initial prominence and eyeballs on your ads.