Glad you enjoyed the piece! A couple other strategies I've found useful are:

1. Apply Chekhov's Gun. In the same way that red herrings often derail the game and frustrate players, offhand descriptions of things in the players environment can all to distract. For instance, the GM might give an elaborate description of the furniture in a room when the furniture itself is not a intended as a clue and conveys to information relevant to the investigation. These descriptions can still foster immersion, but it's important carefully to pull focus back to the key information if players become distracted or to focus descriptions only on what will later be important (if there's a gun in the story, it should be used at some point). This video by Dungeon Craft talks more about this principle: https://www.youtube.com/watch?v=As2dzDCvIh0.

2. If a mystery is part of an ongoing RPG campaign, you can plan the conclusion to act as a clue for the next mystery. The GM doesn't need to map out the entire campaign and all the clues of its future mysteries in advance—the players' will forge their own path and disrupt any plans too soon for that. Instead, I find it useful either to (a) plan just the inciting incident and conclusion and outline my broad ideas for the middle or (b) plan the mystery as "dynamic," with the end of the first mystery signaling something changing in the game world that create a new mystery on the spot or changes the situation. E.g., the villain might ennact some new plan. This way, there's not much to plan ahead of time—each new stage is a response to the players' recent choices. I can then create the clue web as discussed above only after the first mystery is mostly concluded.

Hope that helps!

Best,

Sam