I can think of lots of reasons this might not work in your situation and no one has to justify their reasons. But if you can consider it:
You all know how our computers tend to collect things like your garage, that closet, the drawer from hell - pick the closest fitting personal metaphor. 
Aside from the fairness of blaming Steinberg for the state of my garage, that seems most often to be the root cause found in diagnostic reports shared here. (While acknowledging that bugs happen.)
Microsoft, IBM, Redhat, Docker, Pivotal - every vendor that I’ve been a part of negotiating large support contracts with will tell you (as they explain why won’t apply the same SLA’s) that the highest risk situation for weird behavior and a protracted outage is some kind of update to an older existing server. They recommend “start with a fresh image” because there isn’t really an upside in figuring out why that one was weird or whose fault it was. And especially when the word “decades” seems to come up a lot… I wouldn’t want to somehow damage the configuration of the existing one!
I say it not to defend (I have no stake) but as the best advice I know when you can. On the personal side I usually need a push to replace or add a machine anyway. (Don’t ask about my phone.) But its almost always an immediate source of relief and time savings for everything I choose to do with it going forward. Which eases the sting a lot…