Ok. First let me say that as someone with a penchant for building rather large motherships from which to operate independently in deep space, I wholeheartedly approve of a systems overhaul along these conceptual lines. It is in fact a more realistic portrayal of what real spacecraft actually have to deal with as an operational and design issue.
However, I think that if you're going to go with heat generation management as the primary governance principle for design, you should not make such an intuitively understood concept so counter-intuitive for people. In other words, this is a case where realism can actually save everyone involved a lot of grief while still opening up vast potential for complexity of design and actually scaling in a much smoother way.
In reality, even if a reactor on a ship has coolants circulating around it etc. to keep it from melting down, ultimately, that heat must be radiated away from the ship faster than it is created.
Which means surface area. Which is very relevant to the discussion at hand. A quick thought experiment is in order:
Suppose you were to give us options on how to ditch all that heat?
You could still make it so that conduits carry heat away from the core, but give us choices on where they lead. For instance, if you make it so that when conduits are connected directly to the outer hull/armor of your ship, the amount of heat dissipation is based on total surface area of the ship. This makes it so that heat generation/dissipation scales directly with each ship, rather than having to worry about us hitting the boundary of a 'threshold bracket'.
Further, you could make different types of hull and armor radiate that heat away with graded efficiencies.
Basic hull radiates heat most efficiently, while standard armor, being a cera-metal composite, would radiate heat less efficiently, and so on, making 'shield boat/armor boat' differentiation and variation strategies still viable in designing our ships/stations. This principle could also operate both ways if you decide to implement star heat effects, so that heavier armor would protect you more from the effects of a star's heat on the facing side, even though it presents heat dissipation problems for overall operation.
This mechanic would be sufficient for most standard types of ship, and indeed would encourage some very interesting design variations to maximize radiating surface area while minimizing mass for small to medium combat or utility craft. Cooling vane arrays, more organic, spherical or finned design choices and the like. It also obviously encourages people to create interior living spaces for themselves, as they need to create the volume in order to radiate enough heat away to make the ship functional anyway, but this overall mechanic would allow for greater choice in placement and distribution of those interior spaces.
You could also make the rate from this mechanic slower in dissipating heat when used alone, so that with constant firing of a weapon or constant charging of a jump-drive, etc., slow heat buildup can become an issue over time if systems are improperly balanced against surface area. Interestingly, this would also offer us choices and compromises to make for small high-powered, short-engagement craft like fighters, frigates, corvettes and the like. Much as has often been done with the existent power systems, you can run a ship 'hot'.
Meaning that like now, it would still be possible to mount weapons systems that far exceed the rated cooling capacity of a ship of a given size, allowing for differentiation of design of craft meant for short, intense engagements and those built for long engagements or operation, such as a mining vessel or battleship, which must continually fire it's modules over a long period of time.
For compact, high-power, heavily armored or heavily shielded ships however, (think military-grade vessels built to have minimized target profiles in order to make them harder to hit, or deploy-able fleet miners) this purely passive radiating strategy alone would be insufficient, and would be where your suggested cooling strategy would come into play as a significant augmentation to cooling rate, rather than being the sole means of dissipating heat. This would still encourage the creation of at least minimal interior spaces on such vessels as per your system.
It offers simple to understand complexity in design options for beginners and veteran builders alike.
It would allow us a finer control and greater variation over the potential roles for a craft we design, from highly specialized to highly generalized capabilities. Each would require concessions to be made in design, while doing very little to limit variations in our solutions to such design challenges.
Want to have big sails or very textured and/or expansive and spidery, intricate overall design for your ship that just looks cool? That's actually a good thing with heat dissipation being tied to surface area, but you'll be bigger, easier to hit, easier to critically damage by the removal of outer hull or armor thereby removing heat-sink ability, and it makes you slower to maneuver.
Unless we install your suggested specialized coolant chambers, which force compromises of their own as far as interior space/surface area ratios.
More importantly, this still allows you to balance things quite easily by changing heat generation/dissipation values for individual systems and types of outer hull block, offering you finer control over how to do that balancing, while allowing you yet more simple realism for immersion purposes.
For instance, you could make it so that the primary limiter on missile systems or cannons are based on load time as they are to a certain degree now anyway while generating very little heat, whereas beams or rapid-fire cannons and missiles could be fired continuously, but generate a lot of heat because the magazine of the weapon is working harder, or the beams generate a lot of heat with continued operation. This would allow all the variation in weapons combinations and actual practical functionality of them left largely intact while offering that finer control in balancing both for the devs and the players. It also offers you solid grounding for the conceptual framing of explanations concerning how things are balanced for new players to easily grasp the principles of.
It is very easy for a person to understand that in order to ditch more heat, their ship needs to either be more 'puffy', or more spindly and elaborate in design to radiate it away, as well as it being easy to grasp that such would make the ship potentially more vulnerable to harm once the shields go down as a result.
I hope that what I propose has been laid out here in a comprehensible manner.
I believe it would maximize design variation and game balancing options rather simply and intuitively, while simultaneously placing real and finely scaling, knowable limits on what is possible to do with each vessel or station based on size and/or firepower.
Not to mention easier conversion of existing game assets and ship designs. Power Drain modules and drones used in existing combat designs simply become Heat Beams instead, which offer extra cooling for your ship while pointing that heat at the enemy in an effort to overload their cooling capacity. That sort of thing.
In addition to this, if you decide to leave power generation/capacity as a factor to balance against heat generation/dissipation, then all you have to do is make your 'Chambers' into 'Turbines', and you provide us with the option to simultaneously generate extra power for high-end weaponry etc., (negating the need for Auxiliary Power blocks) while rapidly dissipating the heat the modules being powered generate. You could even make them operate on a toggling basis, just like the current Auxiliaries do for emergency combat-rated use of the ship's full capabilities.
Sorry for the long reply, but you did ask for suggestions on refinements for the idea...