The ammunition per se is less protected, but the armour around it is what you ignore. Yes, when you can penetrate the frontal hull armour, then the ammunition can be incinerated and this is more dangerous to the crew than in tanks where all ammunition is stored in separated compartments. But still all countries design MBTs don't put all ammunition in separated compartments. Why? Because the "if" in the first sentence is a very big "if". Likewise an isolated ammunition storage will not work, if the enemy can penetrate the frontal armour and the doors of the ammunition compartment (and these are much less than 50 mm thick). Again an "if" about which nobody really cares.
Regardless of this, if the armour can be penetrated, then the ammunition stored in the tank is not the primary thing you need to worry about.
Yes, and not everyone automatically gets killed once the ammunition cooks off. The Merkava's ammunition containter and the wet storage of Chieftain and Challenger 1 are designed to slow down the ammunition fire and let (at least the unharmed part of) the crew escape the vehicle. In recent Syrian war there were even cases where members of the tank crew escaped from their tanks just moments before the turret blew off.
A KE round will not automatically kill everyone of the crew, but it's very probable that it will kill or at least wound everyone. Even if it didn't, it still would be enough to knock out the tank in most cases, so that the surviving crew members are stuck on the battlefield in the direct view of a tank that is capable of penetrating their armour. That doesn't sound like "we are saved because the ammo didn't detonate". Blow-out panels are primarily designed following a high amount of tank losses to missiles and RPGs, KE rounds are much more lethal by themselves.
And how much penetrations by tank ammunition don't end up incinerating that fuel? 1 out of 100? In Afghanistan an Marder IFV was hit by a RPG-7 and the frontal armour was penetrated - the only time this was reported. Directly after the penetration the diesel fuel in the engine was incinerated and the crew had to abbandon the vehicle. The current JP-8 of the US Army is even more enflammable (it has a lower flashpoint) and there is a lot more than in a Marder IFV.
You could put the fuel tanks in the sponsons, but that would make the tank only worse. First of all then would not fit. The rear sponson section contains important parts which have to be located close to the engine, like the batteries, parts of the cooling system, parts of the pulse jet air cleaner and parts of the fuel system. So overall only parts of the sponsons could be used as fuel tanks, which are too small to contain all the fuel of the frontal fuel tanks.
But if the sponsons were large enough, it would lead to a decreased protection (i.e. the fuel would be very easy too incinerate, because it covered much more of the tank's profile in the frontal arc) and would require more weight (the sponson tanks would be long, but not very tall and wide -> more surface has to be covered by the fuel tanks.
Your deduction is wrong, that's not how logic works. My logic says to things:
1.) If something is exposed, then you need to have a less exposed back-up system.
2.) If there is a competent back-up system, loosing the original system is not that bad.
For example the turret ammunition storage in the Leopard 2 is not a very big loss, because there are still 1.8 times as many rounds stored in the hull. The Abrams with only 6 rounds stored in (a relatively unprotected part of) the hull cannot afford loosing the turret ammunition storage, because 6 rounds are not enough for any combat mission.
The smaller mantlet could have more armour protection because of the smaller size, just like the Leopard 2A5 mantlet vs Leopard 2A4 mantlet. But then again the Abrams mantlet is very thin, not comparable to the mantlet armour blocks of other tanks.
You and me, we don't know the exact protection level of any tank. Still we can see that because of armour thickness and armour technology, the tank's armour should have an estimated protection level. The thin mantlet armour cannot be very well protected.
No. Hunnicutt says "special armor", a term which includes spaced armour, composite armour and even steel laminates. It could be a simple cast armour element with one or two rolled steel plates welded on top of it. There is no space for any reactive armour elements.
You are using a different definition of "reserve stocks". I cannot see how a bunch of Patton tanks can be counted as a "reserve stock" for M1A1 tanks. In the same sense I cannot count a bunch of T-54 tanks as a "reserve stock" for T-72 and T-80 tanks. By your definition all countries had reserve stocks during the Cold War.
