Design flaws are easier to put into groups than good design choices. While everyone hates bad load times, laggy menus and the game crashing, it's much harder to pin down something down positive.
I am going to try, though.
1. Adjustable challenge levels.
Examples: SO2, ToG, FE8
This is something that I find to be a bigger and bigger deal as time goes by. RPGs are about planning and knowledge of mechanics. There is no physical skill involved with knowing clockticks in FFT, whereas you definitely need to have practice and good reflexes to beat Megaman 1. The age of FAQs mean that you can't really rely on puzzles to slow the player down either. Challenge levels gives hardcore fans of the genre something to leap into, easy mode lets casual fans play the game and just enjoy it. This is becoming more common as time goes by in RPGs, though some of the big flagship games of the series (FF and DQ) have not adopted this. SO2 and FE8 had excellent hard modes- SO2's universe mode actually stands up to the twinking you can do in that game, and FE8's hard mode adds some much needed challenge to a fairly easy maingame.
Tales of Graces gets a gold star here. Not only did it have adjustable difficulty levels, but it let you do so from in game. IE if you have a hard fight or one that sucks (IE some of the sidequest duels) you can turn down the challenge level for a fight. Or turn it up if you want to see if hard mode is worth it. Also it was nice that the game increased drops and combo bonuses on hard!
Bad use of this idea: VP1, FE10, LoM
I haven't touched LoM in over a decade, and I still remember randoms having a million HP in no future mode. VP1's challenge modes aren't really challenge modes and are more confusing than anything else. FE10, fuck the loss of the weapon triangle and threat range.
2. Skits/dialog in dungeons/in battle dialog
Example: Tales series, FF13, WA4
This is pretty straightforward. Mix some character work in with your dungeon crawling. FF13 had no towns (So this was required), WA4 was pretty dungeon crawly as well.
Skits are obvious. If you like them, view them. If not, you can skip. Either way, they are a way to break dungeons.
Bad use of this idea: FFT, BoF5
Gameplay overrides story in FFT, and that is perfectly fine. A lot of the later dialog is awkward due to Ramza not having anyone to bounce off of. The sluice really stands out here. BoF5 ended up very plot light because a chunk of the game's story was NG+. Skimping on the maingame plot and then having a bunch of said plot in NG+ only scenes is not good.
3. Money matters
Examples: FF1, DW4, FE7
Games where you get piles of cash thrown on you and it does no good gets boring. FF1's elfland arc gives you a lot of interesting options. Do you pick up LIT2 and run to the Wizards? Do you level up a bunch and grab silver swords? It makes you think and plan around where you spend your money. If play it smart, you can get through just fine with minimal or no leveling, or you can take it slow and grind the marsh cave into a fine paste. This trails off after midgame (much like the game's challenge), but balancing resources before that is a constant worry.
DW4 is pretty similiar there; it is all about what gear you buy and when you buy it. Old school games that have some balance are able to pull off this type of balancing act well. FE7 gets props as well. (Do you sell promotion items or get shiny weapons?). I don't much care for promotion being tied to items, but FE7 did at least offer up something worth considering there.
Bad use: FM4, FF13, 7th Saga
Flipside, being too tight on money leads to failure. Fuck off forever 7S. FF13's entire money system is flawed; the amount of cash you have available varies hugely based on the randoms you're fighting thanks to enemies not dropping gil. (Of course the entire drop/IC system had all kinds of problems, but that is a different story). FM4 needs to drop a little more cash- you just don't have enough to keep your wanzers up to date without some sim farming, which is annoying.
4. Teamwork!
Examples: FM4, Brigandine, SoA
This is mostly an SRPG concept. Team synergy may be a better term here than actual teamwork. It is about how well the team works together- do you need to use all available PC's for the best outcome? Does godmodding with one character (an N1 failing) happen much?
FM4's link system means that not only do you need to keep your team together for maps, but considreing AP/unit placement/weapons are pretty important. You don't just rush the enemies, you have to plan your attacks and take advantage of counter attacks and conserve high AP attacks like bazookas/sniper rifles while using shotguns and fists more often. SoA's SP system encouraged teamwork fairly well. Your SP has to be managed by the entire group while you balance defense. It works.
The Rune Area system in Brig encourages a lot of teamwork as well. You can't just send a single Rune Knight to solo a map. Unit placement, managing rune power, and making sure your units compliment each other is key. IE: While throwing out 12 zombies in a battle is dumb, so is throwing out six angels and nothing else. This breaks down at high level gameplay, but at that point you are steamrolling the game anyway.
Bad use: Ogre Battle, Kartia, Valkyria Chronicals
OB has some fairly deep mechanics for an SNES game and tried to make you use a variety of units and slowly advance through the stages by capturing enemy towns. That idea isn't bad until you realize you need two good units in the game at absolute most: A high alignment lord unit to liberate towns, and a MURDERDEATHKILL unit to kill everything else. Making sure the lord is competent enough to hold the base is also good. Kartia has an subsection of monsters that are completely worthless; you are better off focusing on the couple of human PC's and getting them leveled up. VC sets up a WW1 style game with trench warfare and capture the flag to go with very distinct classes and- oh wait Alicia/your scout of choice just ignored 90% of the enemies on the map and cleared the level. VC2 fixed this at least.