I think IBM blew it on three areas when they bought Lotus for Notes.
1) Neither Lotus nor IBM treated it as a real developer platform. They brought out new features, but never allowed any sort of compilation or locking of applications. So developers could spend person-years on complex, solid apps, and it was all effectively open-source, so the only revenue model was service, and the next level compelling Notes apps never happened.
2) They never really took seriously the email standardization trend and requirement, as pointed out in the article. I can see why they missed it internally because Notes email in many ways was fundamentally better than regular email, with features/capabilities I still miss everyday, but it was a strategic failure that allowed MS to overtake them.
3a) With the web coming online, they completely blew it on pricing. IBM changed to a high-priced server model, charging by processor cores, often ending up with 5-figure costs just to stand up a server when it had been $35, and not even mitigating it for a small user base, so destroying the entire small business market and small-biz VAR market.
3b) Most critically, with web browsers being free, I'm convinced IBM could have owned the world by making the Notes client free or nearly so. Just the solid certificate/ID/PKI security model is vastly better than anything available today. But IBM basically just forced the entire thing into their corporate client structure and eventually sold off the remains in ~2018, IIRC.
IBM bought something that could really have exploded, and instead of adding the resources to make that happen (e.g., cleaning up the UI issues as mentioned in the article, making it more of a platform, pricing as described), they just worked to jam it (badly) into their corporate software offerings.