So about the sudden disappearance of +4 and +5 weapons and armor…
On the one hand I can see the logic given the new allegiance to bounded accuracy, on the other hand given their opposing natures they cancel each other out, and by the time characters reach the level where they “should” have one I’m not certain that the extra +1 or +2 makes that much of a difference in actual play. Let’s look at the way the “official numbers” add up for characters who are in the right level range to have that kind of magic:
+5 or +6 for Proficiency, +4 or +5 for Statistics, and then +3 for Magic. That makes a range of +13 to +15. A Balor or a Pit Fiend only has an AC of 19 (Hitting 70% to 80% of the time), while an Ancient Red Dragon has an AC of 22 (Hitting 55% to 65% of the time) – with a human in +3 Full Plate having an AC of 21 (thus being hit by the Pit Fiend or Balor a mere 65% of the time and the Ancient Red Dragon 80% of the time).
Granting a maximum of a +5 magical bonus only tweaks those hit percentages up or down by 10% – which while certainly what we would call in my profession “statistically significant” I’m not sure I’d say that it is “clinically relevant”. The bonus adds a minor amount of damage to each strike, and… I’m not in the mood to do the stats on the overall statistical impact of hitting an extra 10% of the time but given that the rules for determining Combat Rating assume hitting with the creatures “best everything” for the first three rounds of combat it is pretty clear that 5e isn’t concerned with that level of statistical detail.
Unlike 1e, where there was some real concern over the need for “bonuses to hit” because of the huge range of AC (a 21 point spread), 5e has what? A fifteen point spread? A Terrasque has an AC of 25, so this is a game where literally a goblin with an attack bonus of +4, without recourse to a Critical Hit, can strike an Ancient Red Dragon 15% of the time (in 1e it was only on a natural 20, but without any doubled damage) and hit either the Pit Fiend or the Balor 30% of the time.
Obversely in 5e, that Pit Fiend and Balor are missing the Goblin only on a Natural 1, the same for the Ancient Red Dragon. In 1e the Ancient Red Dragon missed the Goblin 15% of the time, the Balor missed 25% of the time, and the Pit Fiend 10% of the time! The human (Fighter we’ll say) would, in the same level range of 15+, never miss using the canon 1e rules (natural 1’s only autofail on Saving Throws, not To-Hit).
So “increased reliability” (re: bounded accuracy) really means “everyone pretty much hits all the time” no matter who or what you are because the Attack Bonus rapidly outstrips Armor Class and unlike 1e it never really catches up.
All of this also just goes hand in hand with my general eye-rolling at the “no Dexterity modifications for Heavy armor” that I’ve grumbled about before. Frankly, give humans a Dexterity bonus, and monsters either (or both) a Dexterity or Constitution bonus to AC and those fights suddenly became much more interesting!
AC25 for the Balor, AC26 for the Pit Fiend, and AC31 for the Ancient Red Dragon – with the best human probably having an AC in the 26ish range with a maximum of 28.
Alternately, you leave monster AC where it is but give the Pit Fiend and Balor access to a magic weapon or two and the new bonus to AC for humans is relatively quickly cancelled out to a large degree from a statistical (and “game balance”) standpoint. You could actually make a good argument for this by saying that the increased Hit Dice for large creatures (with correspondingly greater Hit Points) reflects this power rather than raw AC. E.g. yeah, they are easier to hit that the best of the tiny, puny humans, but they are also much harder to kill due to sheer mass and size.
So, 5e is certainly simpler than 1e when it comes to To-Hit/Attack Bonus and Armor Class – but I’m highly unconvinced that it is actually better or that some of the fiddly bits that are supposed to enhance game balance actually matter in most fights.