Communicating about communicating is underrated.
Then comes the adage 'under promise over deliver'.
If you send out an update saying you're working on cool X module/technology/idea to solve a problem, and then it turns out you couldn't, it's a bummer across the table.
Even if it succeeds, you don't get much credit for it because everyone 'pitched in' with their ideas/tweaks at the beginning, so now the entire team feels like you were just the implementer of the team's design/strategy.
Call me a lone-wolf cowboy if you will, but it's a plain truth that no amount of 'team-players' can deny.
And if companies do not want to work with such under-communicators, no big deal. Not everyone meshes with everyone else. There's plenty of other candidates in the sea, and plenty of other companies in the sea.
For example, I would say I get too much email so I really don't want more: likewise IMs or other forms of interruption. Like everyone else, I have things I actually need to get done, so constantly fielding incoming communication can become pretty frustrating.
What I do need are timely and relevant updates, not constant updates and over-communication. I prefer dealing with these updates in a structured fashion, whether I'm giving or receiving them - e.g., weekly 1:1s and other scheduled meetings - rather than ad hoc.
Thinking in more general terms, I prefer behaviour and action that is considered rather than reactive. This goes beyond personal preference though: I don't believe it's possible to effectively scale by being reactive or ad hoc in your actions and communications. I also don't believe that people who cannot (or will not) structure their behaviour are suitable for management or leadership in a growing organisation.
Sometimes events require a more reactive approach but I don't believe this should be the norm.
I think the article have a good point, but it is also important to only "overcommunicate" to people who depend on you. Otherwise it's just noise.
A lot of communication (what I'm going to work on next week) falls under promise. So I don't tell what I'm doing until I've already done it. Or when I'm unable to do it.
In other words, I let success or failure be known only after I've made sure of either using my best efforts. This usually takes ~2 weeks of isolated/concentrated effort to find out.
Then as a consultant at Big Tech working with different customers, I found out they are not big fans.
'promise. deliver or not' - is what they seem to prefer.
Obviously none likes shocks. But they also don't like surprises.
Source - got fired despite extra-mile technical achievements.
Over communication is shallow work. We should limit that. There is a useful amount.
That said, I've received feedback from multiple people that they never know how my work is progressing, are surprised when it's completed, and would like more status updates along the way, so I think it's safe to say that my preferences are an anomaly and you should probably follow the article's advice instead.