The pylon mounted engines are generally mounted so that the compressor fan is in front of the wing so if a turbine does fly apart the blades are forward of the wing spar.
Problem is that at least at the time as I understand it turbine blades were the most likely to fail. With Turbofans a compressor blade failing is the absolute worst case because they're huge in comparison with the turbine blades and so have much more energy. That isn't true of early turbojets - and particularly not of the centrifugal compressor jets we're talking about for the Comet and it's analogues. You're applying modern conditions to a much earlier design and trying to argue they make it wrong - but they don't apply at all.
Podded engines also mean a simpler wing structure. With the engine mounted in the wing the main wing spars have to work around the engine. this causes additional structure (weight). With the pods the main wing spars can be continious. This is simpler to design and build as well as generally being lighter.
Umm... no. There are occasions when the wing spar goes around the engine with no benefits (Blackburn Buccaneer, I'm looking at you), but they're rare. Normally by going to a thicker wing with engines in it you actually save weight - for the same mass of material an I-beam will always be much, much stronger than a simple rod. By punching a hole through the web of the I-beam for the air intakes (putting the engine between the spars makes getting it in and out easier - again not really the case on a Bucc) you get to take full advantage of the extra depth of wing available. It's structural engineering 101, and something everybody was well aware of. Boeing were going all out for thin wings mostly because that was the company design style post B-17 but also because it makes the aerodynamics a lot simpler (you can stick with a single sweep angle and fairly constant chord). If you're willing to go for a more complex and advanced wing planform, you can get all the structural benefits of a thicker wing with very few of the aerodynamic costs.
De Havilland were perfectly aware of both pressure vessels and the stress concentrations at a square corner. This is why the Comet 1 windows were not square. The corners were carefully curved to avoid such stress concentrations. What nobody knew about was the metal fatigue in thin aluminium pressure vessels (ie an aircraft pressure cabin) from repeated pressure reversals (ie climbing and descending) and the incredibly fast tear propagation in such materials when it happens. Knowledge at the time suggested that the curved corners would prevent cracks and, should they begin, routine inspection should identify them before they cause problems and the cracks stopped and reinforced.
Depends how curved - DH went for a fairly square corner, they just rounded off the point.
Worse, they used countersunk screw holes near it which added further stress.
This is the penalty for being the first. Boeing and Douglas immediately reviewed their plans in the light of the new knowledge and later Comets had no problems at all. At the time of design there was no reason known to doubt the chosen design.
Hmmm.... not so sure about that - DH had a bit of a track record of running with very thin safety margins on a design to get stellar performance. The knowledge about fatigue, stresses, etc. was all there - they just failed to apply it effectively. After they were made an example of, everybody else made sure not to make that particular mistake again.
On a different tack; the expected routes for the Comet included fairly short runways so De Havilland went for a low wing loading, thus a large wing so the thickness, in the broad chord Comet wing, allowed for the buried engines with no problem
Ummm... sorta-kinda. Buried engines also allow for high-lift devices along the entire trailing edge of the wing (one reason the VC-10 had tail-mounted engines - it was intended for many of the same routes with short, hot and high runways on them). Problem is, if you look at the safety record of the Comet in particular when operating from those runways it's appalling - G-ALYZ and CF-CUN were both lost to that before they added wing fences and a leading edge droop (and presumably warned pilots of the problem too). I think a large part of the problem is that it was simply underpowered as a consequence of using early engines - later versions had double the power for only a 50% weight increase.
