I think that's absolutely fair. The pace at which the industry is starting to solve really big, hairy problems in standard ways is ever accelerating; the devops space right now is undergoing massive solidification under the Cloud and Kubernetes. UI is one of those problems I hope we're humble enough to admit that we got horribly wrong with HTML/CSS, and maybe we need to go back to square one.
But at the same time, lets pick on Slack. There's no good reason that it needs to take more than 5 seconds per team to load. Maybe its Electron, HTML, and CSS. Maybe they're loading too much data at the start. Maybe their Ruby servers are a little slow.
All of that could be true, but one thing is definitely true: There's a Product Manager or an Engineering Lead somewhere in San Francisco who looked at what their engineers had built and said "Yup, this lives up to the quality our users expect, ship it."
Jira is the same way. Its dog slow. Its full of UI bugs and inconsistencies. But people still buy it, and someone at Atlassian has to have said "we'll worry about this UI bug later, we have more reports for middle management we want to add to the product."
Our best minds will solve the tremendously difficult UI problem one day. But today, the best thing most of us can focus on is the People problem. Expect and Pay for better. Reach out to leaders at these companies on Twitter. A core problem in the software purchasing process is that, often, the people who buy the software aren't the people who have to use it day-in day-out, and often there's obvious lost productivity between the feature bullets on the marketing page and an engineer on an old PC her company won't upgrade for another year.