Quoted from pinball_keefer:

I've been following this thread for some time now and have been fairly fascinated with it. I keep wanting to contribute but never have enough time to sit down and write out all my thoughts. This will be somewhat stream-of-conciousness, so bear with me.
Before I get too involved, just let me state that I don't think WH2O has much strategy at all in that there's only ONE strategy... Get 5x playfield with multiball. If you set it up right and get maximum 5x time, you'll get maybe 6 5x jackpots, which is a total of 15x extra "x" of jackpot value, for what I'm guessing to be an average value of around 300M (at 20M jackpots). If you try and figure what else you can 5x (hazards/camera craze, rafts, Wet Willies (almost impossible), or bigfoot), nothing else is really going to come close to multiball. Maybe bigfoot. Probably not getting more then 5 shots, though, which is 200M. So this just becomes a matter of execution, not strategy.
Similarly, BSD only has one strategy that makes any sense: triple multiball and (double) castle jackpots. Even trifling with bats and rats isn't worth the worry compared to a 480M+ castle jackpot. So again, that is more down to execution that strategy.
On RGP quite awhile ago, I opined that there were 3 elements to what makes games good/fun. Breadth, depth, and (lack of) work. Breadth meaning the number of things that were immediately available to people, depth meaning how many things were "further down the road," and work meaning how much of the game is simply keep making shots for no reason other than the programmer was too lazy to figure out how to make it more interesting (IMNSHO a distinct flaw with most recent Stern games). Breadth and depth generally increase fun, and work takes it away. So, you could come up with a rough "fun" equation of approximately: fun = breadth * depth / work.
I've mulled it over for some time, usually any time I see a "depth" conversation come up, and one thing is clear: 3 scales are definitely not enough. Or if they are, they need to be broken down further.
PinballHelp (I believe it was) was the closest with his river analogy. Most people these days seem to equate depth with "how far until the end of the game," but in the modern parlance of game development, that's not quite the right definition to apply to depth. It's more like distance. How long can the game keep throwing new stuff at you? Breadth is still there, how much is available right off the bat, or to put it another way, how many different things are there to attack in the game at once? Complexity (or interoperability) is the factor most often missing (and the thing that's closest to the modern definition of gameplay depth). This is where the magic comes in, of how things stack with each other or combine in ways to make things really interesting. I would call "depth" some combination of distance and complexity, personally, but I don't know that I can call depth simply how far out the game is or how interoperable the game is.
Let's look at some examples (including my takes on a couple already presented):
STTNG: OK breadth (modes, NZ awards, warp awards, 1 multiball, right ramp awards), below average distance (only 7 modes til Final Frontier, and only 8 total shots to warp 9), medium complexity (modes increasing FF a big plus, but otherwise almost no other stacking possible a big minus), and not much work at all (that is good).
TZ: Great breadth (door modes/awards, camera awards, multiball, gumball, dead ends, hitchhikers, robots, spiral awards), good distance (14 doors, 8 cameras, BTP & EB off spirals. LOOOONG way away for super robots (see also: work)), great complexity (almost anything works with anything), generally low work (getting super robots is a chore, but certainly not necessary; LITZ can be as much of a chore as you make it, personally I prefer to play the game than just go for LITZ).
WH2O: Decent breadth (multiball, hazards/rafts, whirlpool, bigfoot, lost mine), ok distance (6x and WW is a ways away, 6 boulders), very low complexity (5x only stacks with a couple things, and only if you start it first, almost nothing else stacks with anything else), above average work (RIVER 6x and WW is just a lot of shots to make).
CFTBL: Below average breadth (multiball, right ramp awards, right shot awards, MYC), OK distance (multiball has very good distance, right ramp has high distance/work), non-existent complexity (nothing works with anything), medium work (to get super scoring, minimum 12 ramp shots away, often severely higher).
IJ4: Poor breadth (4 multiballs, 2x scoring), high distance (4 of each type of multiball), very low complexity (2x + what), EXTREME work (that's all this game is).
Anyway those are some thoughts off the top of my head that I hope gives you a little more to think about. Great thread for sure.