This is what I’ve pulled together from a combination of my own ideas and talking to a lot of other people about the upcoming changes, hope it helps!
Overall, I’d say the charms are a great idea to open the game up beyond ‘lightning & hammerdin are best at just about everything’ and community consensus seems (mostly) in favour of them as a general concept. But there are certain things or certain interactions that are definitely problematic and could do with some tweaking…
1 - Interactions such as Charm + Infinity / Cold Mastery
2 - Widening gap between caster and melee
3 - Market saturation
4 - Power creep
5 - Making the game too easy
1 - For those not aware, after the Sunder effect makes minus resistance effects apply AFTER the Sunder and apply at full effectiveness. Take a mob with 110% lightning immunity. Currently, Infinity (a lvl 12 Conviction aura ie -85% enemy resistance) works at 1/5 effectiveness against immunities, meaning the mob is brought down to 93 lightning resist so you only deal 7% damage. Other minus resist effects such as Griffons and Facets can bring that 93 down further and significantly increase your damage as a result. Sundered charms change this. The charm brings them down to 95% and then Conviction applies at full effectiveness, so just Charm + Infinity merc, without anything like Griffons or Flickering Flame for lightning / fire respectively, leave the previously immune mob at just TEN resistance. That’s insanely powerful and will catapult things like lightning sorcs even higher than they already are. A commonly suggested solution to this is to make it so that Conviction, Infinity and Cold Mastery only work at 1/5 effectiveness vs Sundered immunes. This is a good idea, but I’d add something else to it and have had very positive reactions to this idea. What if any mob that has had its immunity sundered cannot have the resistance lowered below (for example) 50%? So the charm brings it to 95, Infinity merc at 1/5 brings it further to 78, and then it’s still very desirable to hunt down those best in slot items such as Griffons / Flickering Flame / etc to push the mob towards that resistance cap of 50. This means that your still breaking the immunity, you can still deal damage to these mobs which really opens up a lot more of the game to builds such as fire-based or cold-based ones, but the overpowered interactions taking the resists of immune mobs well into the negatives are no longer a problem and no matter what gear you stack you will still have mobs that resist you will take longer to kill especially on higher player counts, they just won’t be totally immune to you.
2 - Widening gap between caster and melee. Melee / physical builds generally do not benefit from these charms as much as most casters. This is partly because physical immunities are rarer so there’s less need for it, partly because there’s basically no way to reduce the physical resistance below 95 after Sundering the immunity other than decrep / amp dmg (there’s no such thing as physical Griffons or physical Facets or whatever), and partly because %DR on items is quite rare so it’s a lot harder to compensate for the penalty incurred from the charm. I see two main solutions to this. One is obvious and is one that has unsurprisingly been suggested a lot - reduce the penalty so physical builds don’t suffer as much from equipping it. The other is to add physical pierce to the charm. Let’s say the charm had minus 20 to ENEMY physical resistance on it - this would not only mean that sundered immunities are brought down to 75 rather than 95, but would also mean physical builds deal more damage to non-immunes to narrow the gap between caster and melee a bit (they’d just have to make sure in the coding this didn’t accidentally affect PvP!)
3 - Market saturation. Once you find one of these charms for a specific damage type, that’s it. You have it. You pop it in your shared stash if you have multiple characters that might use it and you never need to find another one of that type again, so any more you find just go up for trade and the market probably gets pretty full pretty quick (unless they’re SO rare that it takes ages for a decent amount of them to be found, which would kind of defeats the point of them existing). The solution to this is very straightforward - add a variable roll. Instead of the elemental ones giving a flat -75 to a reisst it could be variable from something like -125 to -75 (just an example), so even after finding one people will still want to find a BETTER one like with torch / anni / skillers with life / etc. Then at the start of the they ladder they’d probably all be pretty expensive, but as time goes on the bad rolls would drop in value and the good rolls would stay pricey, rather than them all falling massively in price.
4 - Power creep. The main part of power creep that’s concerning is the strongest builds getting stronger. If a lightning sorc or hammerdin is currently at 10/10 power level while a fire druid / cold bowazon / etc are at 4 or 5 out of 10 and those builds get boosted to 7 or 8 out of 10, that’s not really a problem because the maximum power level in the game isn’t changing. But if lightning sorc gets boosted to 12/10, THAT’S problematic power creep. The solution in part 1 (making certain interactions work at 1/5 and putting a hard cap on how low a sundered resistance can be reduced to) honestly deals with this pretty well because balancing the Sunder + Infinity / Sunder + Cold Mastery interactions stops those already strong builds getting too much of a boost, but if needed there’s always the option of nerfing the absolute top tiers. A lightning sorc, hammerdin or javazon that loses ~20% of the damage it currently has would still be VERY strong, moderate nerfs wouldn’t come close to killing the builds or ruining the fun, and this helps narrow the gap between the best builds and the worst builds without overbuffing everything else with actual power creep. You can’t ‘just’ have buffs if you want to avoid power creep, if you don’t tone some things down then all you get is things going up.
5 - Making the game to easy. This complaint is more prevalent than seems to be necessary, but it’s only fair to address it. People claim this makes the game too easy, but the bottom line is that Diablo endgame is not hard, there is no inherently high difficulty to the process. Killing Meph 1000 times on a blizzard sorc isn’t hard, killing cows 1000 times on a javazon isn’t hard, walking around on an auradin not even attacking anything isn’t hard. The hard part of the game in terms of ‘requires skill to be good at it’ are PvP, solo-self-found, and speedrunning - none of which are affected. But if more challenge is wanted, and I’d definitely be down for this, there’s a simple fix - make terror zones harder than they appeared to be in the testing people have done so far.
Those are the main things I’ve observed in looking around at what other people have mainly been talking about, with a couple of bits of my own thrown in after getting some good feedback in other forums that I’ve suggested them in. Very much looking forward to 2.5 regardless!!