Except, in this case, it's more like "some people might not prepare for the group project, or bring less than the most fitting ideas forward, so let's reduce our goals for the assignment as not to need anyone to prepare," ignoring that what can thereby be gained is similarly reduced.
________________________
To be clear, I'm fine with cross-class skills being gone. They weren't cohesive with job choice -- instead merely feeling tacked on. I'll admit, there were some enjoyably RPG elements to it, and I liked that it gave reason to sample other classes, but those benefits could be had with less systems-based annoyance.
That said, the problems and toxicity of having acquired Swiftcast (via a lv26 THM) is still as about as deeply felt today for, say, just playing WHM or MCH -- or, until the most recent balancing patch, a NIN, SMN, PLD, or DRG. People still foamed at the mouth over people playing Ad Hoc Samurai in ShB even during the time that playstyle's performance still typically outranked that of the next best melee DPS job. Where do we draw the line on what is reasonable?
________________________
As for the question of balance and diversity, let us consider two games by way of example:In which game is it more reasonable to say "Play what you like; there will always be a use for the class in the long run"?
- The first has diverse fights for which some classes are better than for certain parts and/or functions. Over a given period of gearing/progression, it's not uncommon that certain compositions would be favored for certain fights over others. However, the preferred compositions are far from stagnant; specific interest being made towards ensuring that (aside from the "meta" perhaps favoring adaptable but non-optimal comfort picks that are easier to coordinate in PuGs) a job most advantaged in any one fight of a tier is not most advantaged across that whole tier on average. Class depth, moreover, is sufficient that players can rarely switch from playing their few chosen classes to playing others at the highest level, meaning that there's an opportunity (learning) cost to swapping to the meta class for a given fight, with masters frequently changing what's considered most advantaged in a given fight anyways. Performance is highly varied with the given encounter, composition, and player skill, but a 'lower-performing' job --though still filling a vital function, and perhaps more reliably or responsively than an optimized choice-- can see easily see large (20-33%) differences in throughput.
- The second has procedurally very similar fights as not to shake up class performance or thereby favor any particular jobs over others. The jobs, moreover, play very similarly, especially within their given roles, in order to keep their performance as tight as possible (within 15% from max to min, or within ~7% when excluding the very best and very worst). However, the favored jobs do not tend to change often; their performance advantage may merely shrink or increase over particular patches or tiers.