I suspect the OP may have just discovered that chance is lumpy custard. Which is to say that random events like crank fractures or falling off your bike are not smoothly and evenly distributed through time -because they are random. This means that clusters may occur for no particular reason. Of course, sometimes there may be a common factor, but it’s at least even odds there won’t be.
As to stresses on cranks, like some cyclists I use pedal extenders to place the pedals more naturally under my feet. These add an extra moment of leverage to the strain on the pedal threads. Never been an issue. On the other hand, I once bought a good quality alloy crankset second hand which had professionally shortened cranks. After six months it failed. Not where the new pedal threads had been drilled, which is where I was looking for days to try and find the source of the clicking - it was the spider that failed!
From which I conclude that Chance is lumpy custard.
Chance is a term from the field statistics.
I have had 8 bikes
Fact is that my previous bikes didn't have this tension variation.
Elimination narrowed it down to either the cranks spider offcenter either its mount on the axle offcenter. A square taper connection typical deforms, and that's its very goal: to eliminate wiggle room by tensioning along a tapered path, with the connected parts adapting to eachother. This deformation can bring the spider mount offcenter.
during a period of 30 years.
15 years, 5 bikes, multiple-gears (4 external, 1 internal), but nearly never used them, I just kept on pedaling the biggest front and smallest rear cog, and when the latter became worn, I switched to the 2nd smallest.
11 years, 2 bikes, with singlespeed.
4 years, same 2 bikes, converted to fixed gear.
Over this period, I never had a broken crank.
The 15 years singlespeed and singlespeed-fixed, no derailer, chain tensioned along moving the rear wheel, I never had a chain tension variation - I could just loose 2 nuts, slide wheel backwards, fasten 2 nuts.
Then bike 8, costing 3 times the price of aboves bikes, the first 3/32" stainless steel chainring and the KMC Z1RB wore down in a single month, tensioning range at its end, and a huge tension variation going from 1 to 5 cm up and down.
Crankset replaced to make possible a 1/8" chainring, and chain replaced with a very wide one (3/16" plates).
This setup held out 18 months, made possible only by swapping / flipping cogs and chain, until the final discovery of the reason for the excessive wear (at a moment that heavy chain hung tilted 45°): a 5 mm wrong chainline.
In order to allow a 48T front cog, an even longer axle had to be used.
And this was just a selection of the many more problems that passed by (and were solved).
So, I conclude something else: those two crank failures in less than a month, were not bad luck/chance. I say that there must have been a specific cause. Just alike all before had.
Clearly, the dealer promised a bike he later on found out that he couldn't build to meet the demands.
This whole story, is consequence.
This chain tension variation is not the first problem, it's the last. The other problems got solved, and since previous bikes proved possible, this offcenter somewhere must have a solution too.
That is what this is about. The cranks broke - the alu thread stripped due to cracks that developed from the cranks pedal eye. Clearly some extraordinary stress takes place there. Stress not present on my previous bikes. Steel cranks would be more resistent, but they would be dealing with consequences, not causes. So it's better to hunt, and dump, the cause.
And therefore I decided to dump as much I could from the bottom bracket, to replace it with setups that proved (me) to be better. Octalink instead of square taper.
