Maybe 'RAW,' but it doesn't take much to shovel the grit onto 3.5 - ban PC classes (or caster classes), for instance. OK, it's a major change, but it's an easy one. Don't follow wealth by level guidelines, have the PCs scrabbling for silver through level 10. Should get plenty gritty.
4e it's not so simple, because there aren't a lot of ready-made 'gimp options' for PCs, and less dependence on magic and wealth for basic things like healing. Dark Sun manages some grit with 'survival days,' and so forth, but it's still awfully 'heroic' grit...
You could always re-skin things in 4e to have more gritty flavor, but building it into the mechanics would take more work than banning a few classes and giving out a pittance for treasure.
Somewhere, I have a notebook with a few tweaks I used which seemed to work. When I find it, I'll post the details. Off the cuff, most of it revolved around changing how Solos and Elites work. I had a category of monster which I used as an 'elite' which had the HP and defenses of a normal monster, but with the benefits of being an elite (i.e. actions points and maybe a special trick or two.) Likewise, I did the same with Solos. They had HP and such consistent with an elite, but the benefits of a solo.
My reason for doing this was to reduce grind. However, I also found that I felt more comfortable adding more moving pieces to encounters which involved 'solos' and 'elites' since they didn't take up such a huge chunk of encounter XP budget. As such, I rewrote the XP guidelines to allow for more creatures in each encounter. I also used more minions than the guidelines state. (The minion thing did later become errata I believe, but I don't remember.)
Another thing I did was rework a lot of the math the game world is built with. RAW 4E -in my experience- makes the world too fragile compared to the numbers which PCs can regularly generate. In contrast, some of the obstacles which are easy for PCs are very difficult for monsters. I'm still not sure I got things right here, but it seemed to sorta work. I may have done more work on it had 4E had a longer lifespan.
Another area I started to do work on but didn't finish was saving throws. I always felt saves were too binary. I feel the disease/condition track should have been used for more things in 4E. I was working on a way for wounds, save or suck effects, and various other things to work on a condition track. A failed save, and the condition worsens. Once I started tinkering with that, I looked at using a similar idea for crafting.
Say a weapon had the broken condition because of being sundered. A PC could roll what would essential be a heal check to aid the weapon (with the skill used depending upon the item in question.) On a good roll, the item would improve on the condition track. If it were to get worse, you would need to seek out a blacksmith more skilled than you to get it fixed/crafted/whatever.
Taking a broader view, it seems to me that skill challenges are more-or-less an extended condition track. In this area, I had two influences: the disease track of D&D 4E and 'contests' from GURPS 4E. Instead of skill challenges being so yes/no, there are degrees of failure and success. This may mean there's no set limit to the amount of rolls for something. That is to say there's no set number of X good rolls before Y. Instead, each roll is given meaning. Depending on how the condition of the situation changes, it may be that no more rolls are possible at that point in time.
While none of those things looks especially gritty, I did feel it gave me a little more granularity. With that granularity, I started to fine tune things more. Looking at how the 5E math progresses so far, sometimes I'm curious if the designers were looking at a post I once made on the WoTC boards and just tweaked the idea slightly. Very doubtful, but sometimes I wonder.
