Accuracy comment from the guy at ground zero...
The primary complaint appears to be that Microsoft re-implemented much of the Linux kernel's existing FAT driver rather than building exFAT on top of that driver.
s/Microsoft/Samsung/
There's also technical issues involved - the code started off life as an out-of-tree driver, and it's
really hard to build "out of tree" on top of an in-tree driver.
Now that the code is headed for in-tree status, there's a lot more flexibility to where to go with it. Some people want it to mount exfat only, others want something that will do fat16/fat32/exfat, and it's possible that once the code is cleaned up more, it becomes more obvious how to add the equivalent functionality to the current in-tree vfat driver.
I'm not married to any particular approach, as long as it results in a stable in-tree exfat driver.
(And if you think the code is ugly now, you should have seen it as I got it. I think I'm now at 1,500 lines of code heaved over the side entirely, another 3,000 or so restructured and relocated. It's gone from 28 source files down to 10.