Right. Note your use of the term "lead to". It is a result of a mere bit distribution center pretending to be an OS developer.
No. But NT & Co. have a tendency to become so seriously wedged that blowing it all away and...
That is because this is the only, I repeat the ONLY, option they have. As time went on, those who could have been trained to troubleshoot while administering an MS system farm (especially self-trained) went on to less boring and less frustrating work. What is left are those who would normally have been promoted to PHBhood.
It is a daily operation only because that is the only thing that works the fastest. THINK about that. Now think about it some. To get a computing service back on-line for your customers, the only action is to reinstall. As more and more fixes come into existence, this will take longer and longer and longer until the only solution to delivering this service is to round-robin your users to the last system you've installed; eventually reinstalls are going to take days.
Ms is in seriously trouble because it has not included its customers to also have access to self-training techniques. A large part of their overall goal is to centralize all this know-how within its walls because they have the mistaken belief that this will bring in revenue. It may for the first (guesstimate)50,000 consumers; but they will not have the capacity to service a billion. The reason for this is physical; c (speed of EMF in a vacuum) will never get larger.
It's the "indexing" and "background fragmentation" that do it. M'Snot-NT appears to periodically "optimise" hard drive layout by shuffling stuff around. I built a testbed machine with Lose-2000...