There have been some concerns regarding Cain's proposed Wizard magic system. Mainly that it means we can use every spell repeadetly in longer combat encounters. Can you calm our fears?
I think it depends on what you consider "longer". Practically speaking, even a mid-level mage would have to be in an encounter for a long time to both exhaust entire levels of spells and then continue the battle long enough to see those levels unlock again.
Let's assume a 10th level PE wizard has the same number of "castings" per level to cast as a 10th level wizard in D&D 3E. I don't know how we would want to roll over levels of spells from per-rest resources to timed lockout resources, but for now let's say that the wizard's 5th and 4th level spells are per-rest resources (just like normal D&D) and that 3rd, 2nd, and 1st level spells are on timed lockouts. For simplicity, let's also assume that it takes about the same amount of time to cast these spells in a full round as it would in an IE game. Though we will not use the same timing as the IE games, it's likely that wizard spells will be among the more time-consuming actions to perform.
Using D&D spells for this example, the wizard could cast fireball three times or fireball once, then haste, then slow, or two hastes and a fireball -- in any combination, the wizard has exhausted all three of his or her 3rd level spell slots. All level 3 spells are now locked out for 30(ish) seconds. The wizard would have to cast another five spells before the level 3 spells were available for use again. Either the wizard is going to use up a lot of 2nd and 1st level castings (possibly locking out one of those two levels in the process) or is going to be eating into his or her per-rest resources.
There are a number of tools we can use to balance how this works: 1) the number of castings before a level is locked out 2) the time that an individual level is locked out before it can be used again 3) when a wizard's spell levels roll over from being per-rest to timed lockout (e.g. in this example, maybe 3rd level spells should still be per-rest, but 2nd and 1st are timed lockout) and 4) the power of the individual spells.
Our goal is to allow casters to contribute to combat in a way that is more substantive than hucking sling stones without always needing to chew into a per-rest resource. Additionally, we want the caster's higher level spells to be reasonably powerful but also a strategic resource for the player to manage.