One of my cameras are acting weird, whenever it tries to write to a particular SD Card, the files are corrupted and causes the camera to crash. The other files are still readable, luckily. Is the SD card already worn out from repetitive writes?

Sounds like the SD card is probably shot, they're cheap so just get a new one.

I've went through a few XD cards myself.
Always starts with write errors/corrupted files.

One of the great disadventages of flash memory (One of the reason why I stay away from SSDs) is that they are limited amount of writing times. When all the area of the card is consumed, writing problem may arises.

SSD's are made by far more reliable than "disposable" flash devices

With that said, i only install SSDs at work (IT) or to family (to make the refurbished PCs i made for them work fast). As for me, i use old HD's (which i get for free from trashed PCs). Also, SSDs are far more expensive than HDs for storage space unit

It can be an fake SD card
I had some issues with one genuine sd card: suddenly it went do being an rom.

I've heard that SSD crashes caused by power interruptions is very common, in contrast to mechanical HDD.

I've heard that SSD crashes caused by power interruptions is very common, in contrast to mechanical HDD.

Yep they don't like being interrupted during a write cycle and a mechanical disk has enough momentum while spinning down to even park the heads they are actually designed to do this so they survive surprisingly well. But you can get power loss protected SSD's they basically use DRAM memory and a capacitor to support the DRAM to the large NAND memory writes should there be a power loss, however there is a cost to this so they are not normally found in consumer products, and UPS's are getting cheaper and offer more protection and allow for an orderly shut down and can even keep the desk light running so you have time to find the flashlight, so unless you have critical data, the UPS makes more sense for consumer products, and if you have critical data the UPS would already be there to support the generator turn on delays, and you would have enough fuel to support the mag tape write times which will keep your data for fifty years or more..   

I've run into some compatibility issues between certain Canon cameras and Kingston flash cards.  Often that can be resolved by formatting the card in the camera.  Usually the owner's manual of the camera will recommend formatting any new card to be used with a given camera in the camera.

My suggestion in this case would be to move everything that can be recovered off of the affected memory card, format it in the camera, using the camera's format option in the setup menu, than see how it works.  If that doesn't fix it, drop a couple of $ on a new card.

I have run into a compatibility problem with my SD card... Well it is not a firmware compatibility problem - The card is a micron too thin and the camera contacts are a micron too loose, together it does not recognize the card. Solved with a layer of electrical tape on the back of the card to make it thicker so it presses better against the contacts

Well, the reliability and limited write cycles are different things.
Even when the SSD's have limited write cycles (hence limited life), it does not mean they are not reliable - that end of life failure could be well predicted in far advance, so related problems with data loss are quite easily avoidable.
But if a mechanical HDD fails mechanically due to some mechanical shock in a portable device, such failure comes without any warning, so the data loss is very likely there.

For the flash memory lifetime:
Practically all products (flash hard drive modules, USB sticks, SD's,...) implement some form of wear leveling technique to limit the write cycle of the most exposed blocks. It is done by remapping the cells with each write cycle, so the wear is spread over more cells (because first failed block means the memory failure). But in order to be effective, it need enough free space to be available for the relocations.
So if the drive become full, the device lifetime get very drastically shortened and so will fail very soon. The reason is the impossibility of the wear leveling mechanism to do its work and so some exposed block is going to run out of its cycle life (so cause the complete device to become useless).

And there is another catch: Mainly on the portable devices the wear leveling algorithm is tailored to some particular file system, mainly for the way, how such file system uses different sections of the virtual space (so with SD's and USB mass storage the wear leveling algorithm differs for space used for the frequently rewritten FAT table, vs the way less accessed data area; the difference is mainly in the required redundancy vs usable capacity reduction trade off). The thing is, the wear leveling does not recognize the file system, it differentiates just based on the accessed sector address and assumes some fixed drive partitioning. And the problem is, when someone installs different file system on such device, or changes the drive partitioning. Then the most changed sectors would be on different locations, where the wear leveling is not that effective, so that part then wears off way sooner.

So the two basic basic rules to get as much wear leveling efficiency, so write cycle life as possible are:
- Don't alter the partitioning, nor the file system
- Keep sufficient free space on the drive all the time

Don't alter the partitioning, nor the file system

How is this supposed to be doable with internal drive modules ? Are they actually only optimized for a default windows install scheme ? Or they have better algorithms ?

Don't alter the partitioning, nor the file system

How is this supposed to be doable with internal drive modules ? Are they actually only optimized for a default windows install scheme ? Or they have better algorithms ?

The thing is, if you know the use pattern of different drive area, you may way better optimize the leveling algorithm (it achieves better improvement of the wear, at less cost of the extra memory space; sometimes the extra space is not accessible for the user, so the problem with full drive does not affect its performance - good example is the FAT on SD cards,...).

When the function, so the use pattern of the areas is not known, you may implement just somehow tuned generic, universal wear leveling scheme. Because it can not be tailored, it is not as effective as the specialized ones, so it either can not offer that much life improvement of the most frequently used areas, or it needs more of the free space (whether that is hidden from the user or not is not that much different). So for the same cost it performs worse than the specialized schemes. But as it does not distinguish, you may repartition these freely without any impact on the leveling performance (it stays all the time the same).
The second scenario is mostly used for the generic market internal drives (when you buy a SATA-SSD as a generic part)

The drives for internal use are either custom developed for a given device, so where the partitioning and file system is known upfront (usually brand computers, including everything: HW + file system; e.g. Android devices, or I would suspect Apple as well). Then the same as for SD apples: Don't alter the partitioning, nor file system. These are maily the encapsulated portable FLASH drives (CF, SD, USB sticks,...), some equipment makers use these mainly for the on board storage (internal memory of Android or Apple phone, many notebooks with pre-installed OS,...)