It is easier to achieve your architectural goals with greenfield development than modifying an existing system. However.. once you have an existing system, why

It is easier to achieve your architectural goals with greenfield development than modifying an existing system. However.. once you have an existing system, why are you going to throw the whole thing out, waste all those man-hours spent to make the thing, discard all the bugfixes you had to sweat over. It is very hard for a large organization to justify making something new. It is the reason IBM had to set up a secret ninja tiger team to make the pc. They tried a few times before, and ended up with really nice systems, far nicer than the pc. but they cost 10,000 dollars, in 1980, you are not gong to conquer the small business market for computers with a 10,000 dollar machine. so they set up a secret team, far away from the rest of the ibm mothership apparatus with the mission to make the cheapest computer possible, and they did, it cost 3,000 dollars had nothing compatible with any other ibm and sold like hotcakes.