The main thing to realize is that it's the timeframe of the disease killing you that is important. Anything that kills humans at a very young age is going to have a very strong evolutionary pressure to be wiped out. That's key. That's also why I said that if we want to increase our average lifespan we could make humans have kids at an older age. That would really increase the evolutionary pressure to eliminate diseases and other detrimental things until the time/age that we have had and raised our kids. Yes I know that there's many reasons why we wouldn't likely take this approach but it helps to illustrate the reason and age at which these diseases start to strike on mass. If a disease strikes us at say age 75 it really has no significant bearing on our evolution because we've already had kids and helped them get to child bearing age and therefore that problem is not really a problem at least for our species to survive.
If too many of us live too long then overall resources will be reduced and there'd be less opportunities for younger humans to take their place. Younger humans with possible helpful mutations are needed to continue allowing humans to be better adapted to conditions.
We are a very successful species and therefore most mutations are bad ones. Why would you want to change a good/successful design right? So bad mutations are generally going to be eliminated. Yes many mutations may not be significantly bad per se but if they have even a small deleterious effect they're likely to be eliminated over time. It's almost always only the 'good' or beneficial mutations that survive and help us adapt via evolution. What is good, will depend on what is needed in that habitat and those survival pressures. If we don't evolve fast enough because say we don't have enough beneficial mutations for nature to select from, then we could or would have become extinct.
I hope my explanation helps clear up at least some of the confusion and makes my argument make more sense.