Just to add, requiring specific classes and having obvious optimal party compositions is not strategy, but in fact promotes lack of strategic creativity, instead simply promoting a "go-down-the-checklist" style of thinking.
Strategic challenge comes from the vague middle-ground between "you must have these specific classes/roles" and "everybody can win no matter what they have."
If you can't understand the nuance there then you are just a typical sad nerd who thinks that only binary possibilities exist.
I think most people asking for a game where to beat the harder battles you need roles A, B, C, D, E, F, G, and H. While classes will fill either 1 or 2 of those rolls. So if you need area of effect damage, you can have a wizard or sorcerer, but a fighter or rogue won't cut it.
But that's still rather simplistic in terms of strategy, and better described as logistics. Like I said a posts ago, I don't think permanent party composition at the beginning of a game really is within the domain of strategy (unless there are complex synergies between classes/characters, or the like). Strategy is not a "closed system" but rather strategy requires one to learn and predict about the mindset and the resources an opponent has. Which is why I said it could be strategically interesting if you could scout out dungeons/encounters before-hand and modify your party composition pre-encounter.
But party creation, as it is in the norm of RPGs, is just not strategically complex because you can't use your strategic skills to plan for 30 hours of a complete variety of gameplay. At best you end up with what you described, a checklist of roles, which is on the simple end of the strategic spectrum. Strategic challenge comes from thoughts that are more complex than "I need a healer" or "I need AOE damage."
I don't mean to deride party creation at the start of a game. But thinking about it, I feel it is more of a simulationist advantage than a gamist.