However, what I have found most interesting is in the moments where I notice that a rule from 3.5 not only didn't age well, but it aged so badly that I basically change the rule and substitute a 5E method instead. So far I've run in to this situation a few times. Damage reduction as a rule is just a pain in the ass, and something that is in many ways potentially invisible to the players. If I as GM tell them, "Your blow seems to have little or no effect," the players need to guess if that means they can't penetrate the DR or there's something more troubling going on (like immunity to the attack). If your whole group is striking and rolling low on the damage against a DR protected creature they could be in for a brutal slog. This is where the much more effective resistance/weakness mechanic of 5E is so much more sensible, and luckily incredibly easy to substitute. It even gives the DM easy descriptive language to use: "You strike your foe and it seems especially weak/resistant against your attack," tells the player all they need to know, and if its an actual immunity that is easy to communicate, too.
We've been sticking to low level play in 3.5 so far (that will change over time), but I already know how insane the stackable vs. non-stackable modifiers get as the game progresses along. Needless to say, the elegance of 5E's design eliminates this issue entirely and seamlessly as well.
As fun as advantage/disadvantage is in 5E I haven't seen any reason to introduce it in 3.5 yet. I am quite comfortable with basic modifiers and DCs working as intended, but the ability to award inspiration is missed.
The way 5E handles magic items is so remarkably different from 3.5 that it bears mentioning. The extremely processual design of 3.5 magic items is in many ways one of the Achilles' heels of the system, as it created the rules and expectation of magic economies, and provided hard rules that sort of demystified the entire process of magic item creation. Nothing short of epic artifacts could not be found at the right level of market, or made by the players, with the rules as presented. In 5E it deliberately eschews this entire affair and only in later 5E do we se some rules creep back in (chiefly because as a result of 3.5 more than anything the notion that every city and town must have a magic item economy had become thoroughly engrained in the genre*).
5E's basic rule on training new skills and languages (pay coin, spend time, get skill) is far superior to the class/level restrictions of 3.5. So while I like granular skills and the skill point method, I much prefer 5E's more organic division on skills, allowing for more realism in how people learn things. It also lets the DM incorporate learning skills over time as a reward without breaking any game balance.
The most noticeable detriment to 3.5 over 5E is that rolling a nonhuman character is a pain in the ass, thanks to the scaling issues being dealt with throgh level adjustments. The 5E method is simply less hassle for the same result.
By coincidence (or not) we are resuming 5E on the next live off-week session, so make of that what you will!
*Sure the notion of a magic item shop existed long before 3E, but the idea that it was part of a mercantile economy with clear rules of commerce and manufacturing, and an almost total absence of mystery was very much a 3rd edition introduction. In 2nd edition and earlier the magic item shop was the weird wizard and strange shop in a large city, surrounded by mystery, usually with a specific and well regarded name attached. But the key idea was that prior to 3E, the magic item shop was a contrivance of the GM and fully in their control, and laden with mystery. The magic item shop of 3E was by contrast well defined and something the players themselves could establish if they were so inclined. It stopped being a DM tool and macguffin and became a player utility.