Every enclosure is a thermal design problem
Why sealing a box makes it a heat problem by definition.
Sealing anything that generates heat inside a closed box removes the free, unplanned path by which warm air would otherwise have drifted away on its own, which means the sealing decision itself, made for entirely different reasons like keeping out rain or dust, has quietly created a heat problem that now has to be solved deliberately or not at all.
Every enclosure is a thermal problem, whether or not anybody designed it as one, and the boxes I found most instructive were never the ones built around a fan and a heatsink from the start, they were the ones sealed first for an entirely unrelated reason and only later discovered, usually by something inside quietly cooking, to have been a thermal design problem the whole time.
What is really going on
Anything that draws electrical current or does mechanical work while confined inside a housing gives off some heat as a byproduct, a motor from friction and electrical resistance, a circuit board from every component on it, a gearbox from the meshing of its own teeth, and in an open, unenclosed arrangement that heat simply drifts away into the surrounding air through ordinary convection, warm air rising and being replaced by cooler air from nearby, with nobody having to plan for it at all. The moment that same heat-generating equipment is placed inside a sealed enclosure, that free and entirely unplanned escape route disappears, because the air trapped inside the box can no longer be replaced by cooler air from outside, and whatever heat continues to be generated has nowhere left to go except into slowly raising the temperature of the sealed air itself and whatever it touches. A box built to keep water and dust out is, by the same physical act of sealing that keeps water and dust out, also a box that keeps heat in, and there is no way to have one property without accepting the other as a direct consequence of it, since both come from exactly the same decision to close every gap the enclosure would otherwise have had.
The zipped-up-coat comparison
A heavy winter coat zipped fully to the chin feels perfectly comfortable standing still in the cold, sealing out the wind exactly as it was designed to. The moment real physical work begins, shovelling snow or hauling something heavy across an icy yard, the body starts generating a genuine amount of its own heat through the effort, and a coat sealed tightly enough to keep the weather out is, by that same seal, keeping every bit of that generated heat trapped against the body as well, turning what felt like sensible protection a few minutes earlier into an uncomfortably hot, damp layer that the wearer eventually has no choice but to unzip, even partway, purely to let the trapped heat begin escaping again. Nothing about the coat's insulation changed between standing still and shovelling snow, only the amount of heat being generated inside it changed, and the coat's sealed design, which asked nothing of it while the body was quiet, suddenly demanded an active decision, unzip it or overheat, the instant there was real heat to get rid of.
Why sealing and ventilation trade against each other directly
A vented enclosure, one left with deliberate gaps or louvres, solves the heat problem largely for free, the same way an open coat does, by letting warm air leave and cooler air replace it continuously without needing any dedicated cooling hardware at all. Sealing that same enclosure for a genuine and often unavoidable reason, keeping rain, dust, insects or contamination out, removes that free ventilation path entirely and converts what used to be a non-problem into a genuine engineering task, since now the only ways left for heat to leave are conduction through the enclosure's own solid walls into the surrounding air, or a deliberately engineered path built specifically to move that heat where the sealing itself no longer allows air to carry it. This is not a flaw in sealing a box, it is an unavoidable trade, and treating it as anything other than a direct, guaranteed consequence of the sealing decision is exactly how an enclosure ends up being specified against an ingress rating with no thought given at all to where the heat generated inside it is actually supposed to go. A specification sheet that lists an ingress rating on one line and a maximum internal power dissipation on another is quietly describing two halves of the same design problem, and treating them as two separate line items to be satisfied independently, rather than one combined constraint that has to be solved together, is a common enough oversight that it deserves to be checked for by name during any review of a new sealed enclosure.
The number that matters here
A fully sealed enclosure rated to resist driving rain and fine dust completely, the kind of ingress rating a genuinely weatherproof outdoor product needs, also blocks the one heat removal path that would otherwise have been entirely free, so a decision made purely to satisfy an ingress requirement can, without anyone intending it, make removing the same amount of internal heat several times harder than an equivalent vented enclosure would have made it.
What follows from this
Choosing to seal an enclosure is therefore never only a decision about keeping the outside world out, it is simultaneously and inseparably a decision about how the heat generated inside it is going to get out instead, whether through a heatsink conducting it to the case, a fan forcing air across an internal path, or simply a generous enough surface area and a low enough internal power to let plain conduction through the walls do the whole job on its own. The rest of this set follows that consequence directly, since a heatsink bolted to a hot component only actually works if it has somewhere useful to send the heat it collects, and a sealed enclosure without a real answer to that question has not solved its thermal problem, it has simply postponed the moment anyone notices it has one.