The proposed functionality of attaching a note to a semantic context (eg. a note attached to a person, popping up in email, LinkedIn, chat, etc.) is well taken,

The proposed functionality of attaching a note to a semantic context (eg. a note attached to a person, popping up in email, LinkedIn, chat, etc.) is well taken, but the spatial aspect hasn't been thought through enough. Attaching a reminder to a door only captures part of the intended context, it very much matters whether you are reminded upon leaving, arriving (or both). In meatspace, we can capture that additional context by placing the note on the inside or outside surface of the door (placing it outside also has the effect of making the note public, but we can ignore that aspect initially). I think that in order to show up in most of (and mostly just in) the relevant contexts, true digital sticky notes would have to be both pretty smart and shoulder-surf your activity, which could be quite invasive. Relatively benign example: should you be reminded of the 5-year-old note you posted for a hotel if you're booking a flight to the city the hotel is in? Does it matter if the note's sentiment was positive or negative? How about your note related to a person who lives in that city? After booking the flight? While making a hotel reservation in that city? How about while reading a Wikipedia article related to the history of that same city? What if the hotel is a historical landmark? What if the person is a historian? All too easily, you could end up wading through a flurry of your own notes, reducing the odds you will notice the one you would want to see, but a false negative ends up negating the utility of the note entirely. You can categorize notes and show or hide whole categories as "layers" over whatever you are doing, but now you are mostly back to having to interact with a silo to deal with that.