I know what COBOL's data types are... I used to program in it in the mid 80's. My grief is not with its data types, but with its verbosity. As near as I can figure, the strongest (and I would suggest only) argument against using something like C++ or virtually any other modern native compiled language in place of it is because one simply has a pathological fear of anything new or different, and a belief bordering on religious zealotry (with often similar levels of refusal to listen to countering views or opinions) that no other language could ever do what COBOL does.
I get it that COBOL works.... but it's just so godawful tedious to actually develop in that I cannot see a good reason to use it today other than it may bring you a decent paycheque because you have an employer that still uses it. There's more to life than money, however... and it's possible to still make a good living programming in modern languages that are nowhere near as painful to use, so I don't think even that argument is a great one.
All programming languages are tedious as fuck to use.
Only when you try to apply them to problems that are too far removed from their own domain of interest. In my experience, COBOL development is tedious even in the very domain for which it is intended - business oriented programming. You can do everything that you can in COBOL in python, for instance, and the resulting work would be just as readable and far less verbose.
COBOL programs lack elegance... they are the epitome of the saying "everything looks like a nail when you have a hammer".
You say that now....
COBOL is not at all hard to learn, and bordering on elementary to simply read and comprehend, but its verbosity makes it extremely tedious to develop in. The ease with which COBOL programs can be understood reasonably well simply by reading their code does not go anywhere nearly far enough to justify this amount of labor when it comes to doing actual development.
COBOL isn't hard to learn...
That much is true, but it is tedious as fuck to use as a programming language.
My point was that experts can teach people what things to look for... that was my point... ideally people will learn about the technology themselves from such people and learn what sort of things they should be looking for when it comes to vulnerabilities.
I'm not suggesting that such education should necessarily be freely given by experts without any compensation, but I don't think it's an unreasonable demand on consumers who don't know how to tell if their devices are secure to put some effort into learning.
Promising costs nothing, it's the delivering that kills you.