I think a lot of promise lies with combining activation with resource management (another currently "new" concept), you know, like the PIPs from DBA-esque games from the 80s.
I'd like to highlight my two favourite systems I've seen recently.
Good = Robotech: Tactics
...may have had the most fiddly minis I've ever had the misfortune to assemble (basically a 1:72 Revell modelling project in 1:285 - i.e. 6mm - scale), but I did like the ideas in their activation system. Basically, it was stock alternate move, and each model got the usual move/shoot action, but you got Command Tokens (=1 per model) which you could spend any way you want. You could attempt to steal an activation, dodge or mitigate incoming fire, move extra range (afterburners) - it gave a complete new layer of resource management and activation choices, elevating a otherwise bland set of rules. And the pool of tokens (kept off table) is easy to track, adding a lot of depth to the game for very little complication/slowing of play.
Good = Lords & Servants
...this medieval game has some unusual mechanics, but I thoroughly approve of their activation system. Each player gets 3d6 of activation "tokens." You can keep up to 6 of them to use in reaction to your opponent in his turn. This adds an interesting decision right off the bat - how many points do I keep to mess with my opponent in his turn? Units then have an activation cost equal to that of the leader and a single mini i.e. a group of 5 AV2 troops + a AV2 leader would cost 4AP to move - 2 for the group and 2 for the leader. Moved individually, they would cost 12AP - using leaders is thus optional but very attractive. These rules create a lot of decisions on how to best to move leaders in order to the most economically employ 'grunts' in a way that is organic rather than "forced." Also, units may make extra activations, but at an increasing cost i.e. the 1st action might cost 1AP, the 2nd action 2AP, and the 3rd action 3AP. So there's another cost-vs-reward to factor in - do I move a second time, at an increased cost? The activation pool and the way Lords & Servants have used it create a lot of decisions for the player.
Meh = Bolt Action
Basically you have a token for each unit, and draw them randomly out of a cup to show who moves next. Basically a randomised "alternate move" that sits somewhere between card-based activation and alternate move. A bit meh. It's got the same tokens as the first two examples, but there's no "resource management" or extra decision points being introduced.
I've been thinking about ways you could add depth to it.
1. You can "set aside" tokens next to a model to store an action (aka overwatch) to react to enemies. Ok, nothing new here. Bolt Action sort of does that already. But what about....
2. You can use these "set aside" tokens to do a group move with a group of minis who are within a certain cohesion distance. This means a group move might take a while to collect and "build up" enough tokens to move all the guys you want to, and enemies could kind of see it coming.
3. A "leader" would allow a group move without buildup. I.e. you only need one token to activate the leader, then you move as many as you want within cohesion distance. Of course, you'd then have to remove some tokens remaining in the cup equal to the extra minis you moved. This makes leaders useful, as they can spontaneously order group actions without waiting to "build up" enough tokens.
Why Don't Actions "Carry Over?"Most rules have a section saying "any unused actions/moves/shots etc are lost and do not carry over until next turn." But why not? Yes, you would need some way to track this. But we're already using tokens on the tabletop from the example above, so let's add:
4. One unused token can be used to give a model an extra action in their next move. So you can "build up" or "store" momentum. This represents models psyching themselves up for a charge, or pausing to collect themselves. That way you could store momentum in a "lull" to spend it later in a frantic attack.
There. The tokens, which were only serving as a random activation, are now clearly a "resource" to be managed. Hmm. Might work for my homebrew medieval rules, actually.
Anyway, those are some samples of resource management-meets-activation, and an example of how random activation tokens like in Bolt Action could be adapted to add an element of resource management to a game.
....I've got some further thoughts on resource management but it's my 2 year old's bedtime*... another day, perhaps. (*Am I the only one who thinks Dr. Seuss was definitely on drugs back in the 60s?)