What are the dangers of CHT going below minimum in flight during descent since the power setting will be around 10inches in a power descent?
To my knowledge, the problem with running low CHT is mainly related to incomplete combustion, leading possibly to fouling of spark plugs and buildup of carbon and lead (if present in fuel) deposits in combustion chamber. How severe this actually is, I have not studied the subject any further, or looked inside an engine during such a condition ;)
However, from experience I can tell that after a prolonged descent, running the engine cool, the engine usually has some "difficulty" picking up the pace as the descent is stopped and power increased. Some sputtering and hesitation are present sometimes, but after a while the situation is normalized. During training, such descents were advised against, and I would say for a reason.
In extreme cases very low CHT might lead to stalling the engine, as it can no longer support combustion due to lack of heat and possibly the fouled spark plugs I mentioned earlier.
It's mostly about the cooling rate. Lycoming recommends limiting cylinder temperature reductions to less than 50 deg F per minute. It's especially critical when cylinders are above 400F. The time when this stress is worst is going from climb power or full power all the way to idle. Power reductions from cruise, if made gradually, generally won't exceed the cooling rate limit if you take it easy, especially if you have cowl flaps to allow regulation of cooling flow.
Once cylinders have cooled down, structurally there's not a problem; the main problem is just the tendency for plugs to foul after a long run at idle. Most flight instructors will tell you to give short bursts of power during the descent to help keep plugs clean and make sure the power will be there when you need it. Also, keep the mixture lean unless you need to go back to climb or max power.
I tend to also bring the power up gently when leveling off because I'm that sort of person, although there isn't really a problem with shock heating, so don't worry about sudden power applications doing damage (still, don't just shove it forward instantly). Lycoming says if the engine can take full power application without stumbling, it's warm enough.
The cooling rate issues are mostly to do with the different expansion/contraction rates of the materials in the cylinder. Effectively, the difference in aluminum (head) and bronze (valve guides) contraction rates vs the steel in the valves and valve seats (piston rings too, but the rings are running in a steel barrel so the effect is minimal).
While heating, the non steel parts expand faster than the steel ones; no problem (the interference fit of the valve seat has to account for this so it's still tight in the head when the head is fully expanded).
While cooling however, if the non-steel parts shrink faster than the steel ones, you get potential valve (mostly exhaust) sticking (can become a serious problem because it can bend push rods) and massive hoop stress in the aluminum casting around the valve seat, as the aluminum shrinks faster than the seat. Cylinders almost always crack between the exhaust valve seat and the spark plug hole, because that little isthmus of aluminum between the two holes is the ideal starter location for a crack.
I have a lot of experience in glider club operations where the engines run full bore, then not, on every flight cycle, and it's murder on cylinders. Clubs in the 90s started to use special power reduction protocols to keep the cylinder cooling rates in limits after glider release, and this drastically reduced the incidence of cracked cylinder heads. This is a big deal going from full throttle to idle; going from cruise power with a lower cylinder temp is not so bad. It was found from the glider club research that once the cylinder head was down to the low 300s F, high cooling rates below that had no significant effect.
So, in general, if you just take it easy on the poor old engine, by reducing power gently at first until the cylinder temps have dropped somewhat, it'll be ok.
Manifold pressure in the descent isn't the issue; maintaining an adequate cylinder head temperature, and reducing the rate of CHT change is the issue.
Keep in mind what manifold pressure is: it's a measure of vacuum, or the lack of it. It's a measure of pressure in the intake manifold. There are a lot of myths floating around about manifold pressure, ranging from the notion that manifold pressure and engine RPM must be be the same or never "oversquare" (very untrue), to my favorite, an engine failure leads to zero manifold pressure (was actually told that by a F22 pilot after he had a power loss in a recip powered airplane--confirmation bias).
Think of the manifold as the hose on a vacuum cleaner. Shut off the vacuum cleaner (shut off the aircraft engine), and the manifold pressure will be barometric pressure, regardless of throttle position (so, 29.92" hg or 1013 mb on a standard day at sea level). Start the engine (start the vacuum cleaner), and close the throttle, and the manifold pressure goes down, because the vacuum (engine) is drawing air from the manifold...if the throttle is closed, it's reduced the amount of airflow through the manifold, so if the engine (vacuum cleaner) is drawing more air than flows through the throttle plate, the manifold shows lower than barometric pressure. That's all it is. The throttle is an airflow valve, and the engine a suction machine.
As a result, for 10" of manifold pressure, one can't say that CHT is low due to manifold pressure low. One needs to look at mixture, cowl flap setting, and airspeed, and other factors, to determine what cylinder head temperature (CHT) will be.
As a general rule, it's not good practice to make idle descents in piston engine airplanes. The common guidance is to keep the power setting "at the bottom of the green," which in most airplanes is about 15" hg. That's generic guidance, with the intent of not only giving enough power to keep the engine warm, but to reduce the propeller driving the engine. It's best if the propeller is producing thrust, or close to zero thrust, but not less than that.
High temperature is a bigger problem. The interface between the steel cylinder barrel and the aluminum cylinder head is what keeps the cylinder head in place. If that head gets too hot, burning valves or warping aren't the only dangers. Lifting a cylinder head, or in other words, separating the cylinder head from the barrel, is a danger. The cylinder head is fit to the cylinder barrel with an interference fit and some slight features on the barrel that help prevent the head from coming off. Mostly, however, it's just a tight fit. Get the cylinder head hot, let it expand, especially if it expands too rapidly (gets heated too rapidly), and the head can separate from the barrel under pressure from the combustion chamber. This invariably results in additional damage and possibly a fire.
Piston engines should be warmed up gradually and cooled gradually; there are aluminum parts, and steel parts which warm and cool at different rates. Aircraft engines are quite tolerant of a wide range of temperatures, and even the effects of air cooling (and cooling from rain and precipitation), but it's best to respect the engine and not abuse it. Consistent operation, and operation at enough power to seal the rings and prevent blow-by, at mixture settings adequate to prevent detonation (not usually a factor below 75% power for a given engine), and of course regular operation within the engine operating limitations, will go a long way toward helping ensure long engine life.
There is both anecdotal, and actual evidence of cylinder head cracking or "shock cooling" from rapid cooling of cylinder heads. I've changed enough of them personally, with cracks in the head, often between the spark plug holes, but also between fins trending toward the barrel interface, to offer anecdotal experiences. It's also accounted in various manufacturer information stretching back before the second world war. That said, there's also a lot of myth surrounding "shock cooling," and some strong arguments made that it's more myth than reality. My preference is to not abuse engines, and that's especially true of high performance engines, geared engines, turbocharged, supercharged, or turbosupercharged engines which can experience some significant temperature ranges and some high power settings (and potential for detonation, another subject). Carry some power in the descent: it's not a turbine engine. If you have cowl flaps, watch the temps, work the engine, airspeed, and cowl flap setting accordingly, and if needed, start your descent a little earlier, with a little more power. Problem solved.