True it takes a long time for formerly great organizations to die. I think a lot of the problems with this type of analysis are time scales and costs of switch

True it takes a long time for formerly great organizations to die. I think a lot of the problems with this type of analysis are time scales and costs of switching. IBM is probably not the best solution for anything if you're starting something new; but if you already have something it's probably good enough and it would cost too much to change right now. So IBM can limp along until new organizations develop products that are sufficiently cheaper or better that it makes switching the logical move. Or the companies tied to IBM eventually fail and replaced by more nimble organizations.