Besides, In flash drives there's a mechanism called wear-leveling,so bigger the driver,the longer the write life..
That is, unfortunately, a misconception. The flash translation layer (FTL) in NAND flash based storage device does indeed level the wear evenly across the NAND flash device (or a section of it), but larger devices don't make for a longer life.
The main reason is that the controller reserves a
percentage of the NAND flash device capacity for wear-leveling (and bad-block replacement). This percentage is determined by the flash drive OEM, usually based on the properties of the actual NAND flash device, but more than often just guessed. This percentage largely determines the life span of the drive. Eventually the reserved blocks are all used to map out bad and worn blocks and the device will start producing I/O errors and/or corrupting your data. Drives with a larger percentage will last longer. So will devices with devices with a larger capacity
and with the same erase block size
and with the same number of erase cycles, but those just don't exist. Controllers with multiple NAND flash chips (or multiple planes within a chip) don't wear-level across chips or planes. Many controllers don't even wear-level across a whole chip, but just across small sections of it. So that doesn't help either.
I tend to say that larger flash drivers wear down quicker than small ones, mostly because they have higher density NAND flash device with (much) larger erase blocks. And high density NAND flash device erase blocks often have (far) fewer erase cycles on them before they wear out. Where erase blocks could once be erased up to 50.000 times in old devices, in modern high density devices this number can be as low as 100.
Also note that the controller is not aware of what area of the flash drive is in use and what area is empty. It just makes its capacity, eg 16GB, available to the user and that area is what it tries to keep working flawlessly. It has no understanding of the file system you put on it, let alone which areas are allocated ('in use') and which are not ('empty').
Despite the vast amount of USB flash drives (and SD-Cards) offered on the market, there is only a hand full of manufacturer of the controller on it. SMI is the largest and the cheapest, Phison and Hyperstone are much better, but more expensive. The controller largely determines the reliability of the drive, but unfortunately 'reliability' isn't a sales argument. Reliability is assumed, even though many drives aren't reliable at all. What is a sales argument is 'speed'. Unfortunately, speed if often gained at the cost of reliability. And to make things worse, the brand name drive you buy today, is very unlikely to have the same NAND flash device, controller or firmware inside it than the identical one you bought last month.
An exception could be SanDisk. SanDisk rolls their own controller, their own firmware and their own NAND flash. And because they do everything themselves, they have better understanding of all these three critical components, allowing them to make more reliable drives (and SD-Cards). From all devices and all controllers I have tested, all SanDisk devices were very reliable and pretty darn power-fail robust. They weren't the best, but they were the best from the non-exotic ones. Unfortunately SanDisk did not want to make any agreement to keep reliability at that level. They didn't tell us, but we assumed they wanted to reserve the right to make their devices faster, at the cost of power-fail robustness. Having said that, they are still pretty damn good, and still the only devices I buy.
