The big problem there is explorers. Any wear and tear fast enough to be noticeable for miners, high-end traders, etc. is going to take them down pretty quickly.
A shielded 5B FSD as usable on many common exploration hulls, with the double-braced experimental, would give about 300 integrity points ... but in practice once you get below about 70% the malfunctions are going to get really annoying, so about 90 integrity points to play with. You'd probably still get about 40 LY range out of it easily enough on an Asp or Krait that was otherwise stripped down, with Guardian FSD Booster ... probably more like 20LY if you've also B-rated most of the other internals so they don't fall apart either.
So 40*90, you'd get about 3600LY before needing repairs, with the ability to maybe struggle on another 7500 LY if you ran out of AFMU and synthesis materials. A standard 5A with long-range mods+experimental as currently fitted has just short of 100 integrity, so 30 jumps before it needs repairing ... not far at all.
That would be a big change to exploring, certainly - pretty much requiring anyone going any distance to carry an AFMU, an SRV, and a mining laser as backup just to keep their ship operational, and potentially top up supplies at every system with a suitable source of raw materials in case they hit a run of systems without. Not in isolation a bad gameplay design, but it would be extremely tricky to introduce now and would need Support to be very generous with the teleports for the "I logged off after DW2 and now that I'm back my ship is falling apart and I can't fix it" cases.
The other problem with high maintenance costs is that it discourages having fun with ships. At 350,000 credits per long-range jump in a Cutter (plus however much all the other modules cost in maintenance, of course):
The result is that Cutters aren't for having fun in, they're for min-maxed money earning and you then fly a Cobra III for having fun.
- that's basically ignorable for miners. Hundreds of millions worth of mined goods can easily absorb the few jumps needed to get to sell up
- it's not a big problem for combat pilots either, who can just settle in a system with good combat opportunities for a while (though other module costs will get them, of course)
- it's definitely on the painful side for traders. With 700t of cargo they need to make a profit of 500 credits per jump per tonne just to break even. That's possible, but rules out a lot of cargo types as uneconomical.
- it adds up really quickly if you're doing something fun but not money-earning like sightseeing Guardian sites, or reading tourist beacons, or going canyon-racing, or just lounging around on a planet surface chatting.
It also incentivises doing everything in the most boring way possible - with trading, you're losing 500cr/tonne/jump if you go fastest route. But if you go economic route and do ten jumps for every one, you only lose maybe 100cr/tonne/jump - so in exchange for seeing the jump screen ten more times, you get more profit. In exchange for not boosting or even raising the throttle over 25% in supercruise or normal space, you take a lot longer to get to the station but save on wear and tear on drives, power distributor, etc. In exchange for driving around for hours refilling raw materials between trips, you can use an AFMU to cover the repairs instead.
I'm not saying that a game about careful micro-optimising resource levels so that they hold together long enough to do the next bit and get more resources can't be fun - just that Elite Dangerous clearly isn't in the survival genre, and it's far too late to move it to it now.
Explorers would be hit pretty hard, as it would turn long-range exploration into an actual survival game rather than simply tourism. Small journeys would be fine though, with a decent FSD a budding explorer could quite easily head out to a nearby nebula such as the Coal Sack, while further afield Nebulae around the 5000ly mark would be reachable with an AFMU or two. Only super-long distance exploration would require the full "living off the land" playstyle, which would add some much needed difficulty to exploration as it is pretty risk free at the moment.
Existing explorers and their stripped-down AFMUless ships would have to be considered, but that could easily be solved by simply only invoking the changes upon docking at a station to ensure that they aren't stranded. Alternatively, they could simply ask every player when they log in whether they wish to be teleported to the nearest station or whether they are happy with their current load out in the black.
The costs I mentioned were actually relating to the cost of outright buying the module divided by the amount of integrity it has, not necessarily the actual cost of repairing it. I just used that metric because the cost of repairing a module scales based on its buying price, so the buying prices can be used for direct comparisons for running costs. Considering how even the rebuy is only 5% of a module's cost, even an increased repair cost would likely be around 2.5% of the buying price to fully repair a broken module. This would put a Cutter down to 8.9K per jump with an A-rated FSD, which is minor on its own but can add up over time as well as the other modules would contribute additional operating costs. This is also assuming that module costs don't get a wholesale rebalance in the future to account for the ongoing income inflation.
Other professions would obviously draw their maintenance costs from different things gradually decaying during use. As you pointed out, combat has plenty of opportunities for a ship to break down over time as it places huge amounts of stress on almost every module other than the FSD; their FSD might emerge unscathed, but an intensive bout in a CZ might leave your thrusters, distributor, weapons and shield (and maybe other modules too) looking pretty worn out even if your shields never go down. Mining would be an issue, as you don't really do anything intensive that would place stress on the ship systems and core mining even places less stress on limpet controllers and the refinery than traditional mining on top of having completely crazy income rates, the only possible avenue of cost adding would be to make the seismic charges themselves very expensive to rearm; either way, mining needs a serious review on the profit it generates either way.
Ships can be used to have fun, particularly the cheaper ones. High operating costs would mean that players would want to use the smaller, cheaper ships to mess around in rather than their main flagships. If you want to go sightseeing or canyon racing, then do it in a DBX or Eagle rather than your Anaconda (plus, who goes canyon racing in a big ship anyway?). It would encourage players to mothball their flagships, letting their other, more efficient ships shine for the normal duties while the flagships get brought out for big events and tasks that a normal ship wouldn't be capable of doing. It's the same in any economic game, you use the cheap and expendable units/vehicles/tech when messing around and save the expensive stuff for actually getting stuff done. It's just like in real life how Destruction Derbys are a thing, but you will note that they always include a full roster of dirt-cheap cars rather than event organisers sending fancy supercars into the ring.
Elite dangerous, in the early days, used to be a game about economic survival. Every action you took had to be carefully considered regarding the risk/reward and, if you played the cards poorly, you could easily end a 6 hour session with less total wealth than you went in with even when you were trying to earn. Nowadays, you pretty much have to actively go out of your way to lose money.