In this question we addressed a problem in which my players were finding combat too easy, and were not having to ration their stronger abilities. After discussion I've decided to implement Dale's suggestion: redefining rest periods to 8 hours (short) and 1 week (long) in order to space the recommended number of combat encounters sensibly for a story-driven campaign while keeping my players' ability use in check.
In most instances I feel this is an optimal solution. The druid I mentioned in that question, for example, will have to be much more judicious in his use of wild shape. His spells will also regenerate at a reduced rate, so the decision of whether or not to use spell will possesses more gravity. That said, this option does come with its own problems... An example of where this breaks down in my mind:
Prepared casters, such as the druid, are typically allowed to choose spells after a long rest (~1/day), making them useful for handling utility spells (such as Shatter) that spontaneous casters would be less willing to learn due to the fact that they wouldn't use them frequently. Requiring a week of downtime to prepare new spells is far from ideal; the party won't have fun camping for a week outside a ruin to allow a caster to prep Dispel Magic in order to break a ward keeping them from entering.
Outside of things that are sensibly altered to fit a more spread-out combat schedule (e.g. wild shape use, spell slot replenishment, heals using hit dice, etc.) what game mechanics may be inadvertently broken by changing the time required for resting?