My general reaction is to ask some version of the following questions: "OK, so let's say this is my tenth play-through. Very approximately, what are the odds I will encounter a foe ("monster") that is significantly different in both depiction (how it is drawn, described, etc.) and in function (how it is endured, contained, defeated, etc.) from anything I have encountered on the previous nine playthroughs?" and "Putting aside explicit foes such as monsters for a moment, what are the odds I will encounter a world situation that pushes a significant shift in priorities or play style from what I have previously encountered?"
Most games don't manage either of the above. A few games do one reasonably well, typically the second (world situation); Civilization or classic ASCII Roguelikes such as Angband for instance. Starting in a mineral-rich desert is quite different from starting in a food-rich floodplain, and either early acquisition of a powerful item (e.g. Thorin's Shield, a powerful artifact granting acid immunity and therefore to approach some situations usually fled from such as jelly pits) or unusually delayed acquisition of a fundamental ability (the dilemma of progressing without free action beyond a certain point) can radically alter not only the tactical decisions, but strategic and even meta-game playstyle.
However, the foes ("units") in Civ are fairly predictable; almost all players will have encountered the full catalog in a few playthroughs. Angband has a much larger catalog and a notion of rarity, so it's possible you might see a new monster; but it would be likely to be a straightforward variant of a type you've already seen.
Central to all of this is the idea that DF is "not just an (adjective) fantasy world game, but an (adjective) fantasy world game *generator*". For some of us, the fact that you currently largely experience the worlds generated via the lens of a small band of cranky alcoholic dwarves with pick-axes is almost a side-effect; a hypothetical "hypochondriac druggie gnomes with crowbars" game, while seeming to be a similar riff, is missing the point entirely.
To try and come up with an extensible analogy, most ordinary computer games are like a restaurant that offers spaghetti. It comes with their house red sauce, you can get it with or without meatballs, and can add cheese and/or pepper flakes. If you ate there for a week, you'd pretty much cover the options, especially if you have strong positive or negative preferences for some of the choices.
A good non-DF game is like a restaurant that offers a "choose your own pasta" setup; perhaps a half-dozen sorts of pasta, stock red or white sauces with some simple variants, and a handful of meat options. You could eat there for perhaps a half-year without an exact repeat, although some variants will end up reasonably similar.
DF is like a pasta restaurant run by an insane chef who owes you a huge debt; it's got two dozen types of pasta, a dozen stock sauces with a half-dozen variants each, two dozen kinds of meat, and a re-purposed salad bar full of add-ins or toppings. Even an ordinary person could eat there for 50 years and not repeat even the base meal, not counting the toppings. But in addition you can just call up the chef and say "You know, I was just thinking I've never had alligator pasta before, why don't you get some in and, hmm, do it up chicken-fried overtop some carbonara for tonight?" and there will be a moment of silence at the other end and the chef will say "For you, anything. Give me until 8pm, though, I gotta send a guy out and then do some prep myself."
Most people who say a game is "like DF" mean "We have more than one type of spaghetti!", as if that began to approach the depth and detail of even the current state of DF, let alone the intended 1.0 game.