A perfect lack of bias is impossible, even in principle, because the DM always has knowledge the players don't. There is never a situation where the players are, on the net, better-informed about the state of play than the DM is; at absolute best, with perfect communication and understanding, there will still be information that wasn't requested or that the DM knowingly keeps hidden for the purpose of suspense. And such perfect communication and understanding is, itself, another impossible goal. Further, things are shaped by the preferences and priorities of the DM. They should be; that's the whole point of having a DM at all. But that very point is what means diamond-absolute lack of bias is unachievable, even undesirable.
Rules, of whatever kind, thus exist to facilitate effective play and consistent results. Humans are notoriously bad at correctly processing probability for example, with slight differences in question presentation easily swaying answers even from well-trained experts in statistics. Yet correct handling of probability is essential for a truly unbiased arbiter in the D&D play space!
Hence, the appropriate function of rules is to help shore up those places where we flawed, imperfect humans are liable to make errors, even when intending to do things right. E.g., if I recall correctly, both 4e and 5e discuss why it's actually very unfair to players to ask them to repeatedly make stealth checks every time they attempt an action--because even though it might seem appropriate, doing that is equivalent to forcing the player to fail eventually. (Even if they only fail 1 time in 10, they have less than a 50% chance to pass 7 checks without any failures.) This is exactly the kind of problem that rules, and especially extensible framework rules (like 4e's Page 42) are ideal for solving, because they provide a consistent, reliable backbone that can eventually be trained until it becomes reflexive, at which point the difference between "Bob says so" and "The rules say so" has vanished, because Bob has made the rules second nature.
And that, unfortunately, is the big problem with most edition changes. Good DMs (and, usually, good players as well) have learned to make the good rules of the old system second nature, and learned to make the bad rules as minimally impactful as possible, or replaced them with house-rules that are not bad. (3e, for example, tended to end up with more of the former than the latter because of how pervasive the system problems are, but it relied heavily on both. 4e tended to rely mostly on the former for the places where it was weak, as outright replacement of the rules was rarely necessary and often done by the designers themselves with errata.)
I'm actually glad of reading and responding to this thread, because teasing this out has finally answered a question I've had for a long time: Why do veteran players of an old system, who can ride roughshod over that system's rules whenever they feel like it, so consistently feel "trapped" or "constrained" or "incapable of X" (whatever X may be) when they play a different system? And, as a corollary, why do systems that go out of their way to be similar to past systems (as with 5e going out of its way to scrub out similarities to 4e even when it uses 4e concepts, while actively playing up similarities to 3e even when the two differ) tend to have less of this kind of response?
It's because a system that feels unfamiliar, no matter how similar it may be in practice, fails to engage that instinct-level incorporation of the rules, and thus the rules feel alien and constraining, even though they don't limit behavior any more than the old rules did (and may even objectively limit it less). Whereas a system that feels familiar, no matter how different it may actually be in practice, preserves that instinct-level rule ingraining, subject to a few tweaks (e.g. 5e doesn't do iterative attacks, you just get more attacks if you have the Extra Attack feature, which feels similar to 3e even though it's arguably more similar to 4e's way of improving Basic Attacks and At-Will powers.)
This is why Mearls (IMO incorrectly, but not without merit) claimed that "mechanics are easy, feel is hard" or something to that effect. Because actually balanced mechanics are MUCH harder than feel-in-the-generic, and that's what I thought he meant (and, frankly, it probably is what he meant). But the actual nugget of truth here is that a mechanic that feels familiar is very hard to make, and that feeling of familiarity is essential to get people on board, even if the feeling is based on nothing particularly mechanical about the rule.
This also goes a long way to explaining why people who did not like 4e often seem deeply confused when 4e fans don't like things in 5e that seem so obviously 4e-like (as I mentioned in a surprisingly popular post in a different thread a few months ago). That is, 4e fans had internalized enough of the rules for, say, Healing Surges that the new 5e rules for Hit Dice do not trigger the feeling of familiarity...but we can actually identify why that feeling isn't present because 4e prioritized transparency in its rules. With systems that do not prioritize it (or, as with 3e and arguably some earlier editions, seemingly try to avoid transparency), all one has is the feeling without a clear origin point unless one is extremely well-versed in both systems' designs, not just instinctually but academically. E.g. the fact that Fighter in 3e felt disappointing was hard to articulate for many, until well-versed players pointed out things like how the magic item loot tables were an anti-transparent Fighter class feature, or how the weight of heavy armor was literally an XP penalty in exchange for a higher survival chance (in systems where GP=XP), and thus Fighters were more able to endure that penalty (and had faster XP tables than Wizards, so a Wizard in heavy armor would be double-extra penalized).