Is there a chance C11 and C12 are switched? In the picture, C11 is physically larger than C12. Electrically, C12 is larger (100 nF vs. 680 pF). Different voltage ratings could explain it, but it still looks suspicious.
Just take a look at the numbers on these two capacitors. 100nF should have printed 104 on it (100000pF), 680pF would be 681 (two digits number + 1 digit exponent).
Small capacitors are often made using C0G which is a type 1 ceramic which has no capacitance over voltage derating but has a lower dielectric constant compared to X7R ceramics typically used in high quality capacitors with higher capacities. X5R, X7R and similar ceramics are type 2 which have a derating of up to 80% capacity loss when biased with a voltage.
So for circuitry where low capacities are needed and linear behaviour is required (as in audio circuits, clock circuitry etc.) C0G capacitors are used which are mechanically much larger than type 2 capacitors with the same capacity.
So the size alone does not say too much about the capacity.
I hope this helps you to determine if the capacitors are swapped accidentially.
I strongly recommend not to buy any standard parts from dubious sources like aliexpress etc. The chances are quite high to receive counterfeits that either don't work at all or don't work to specifications.
When buying parts I always choose well known suppliers like Mouser, DigiKey (if I can order through my employer) or from Conrad, Reichelt, Segor etc. when buying myself (depending on your country there will be other local alternatives.
The last time I had extreme trouble was when I tried to repair my Harman/Kardon Soundsticks. The audio amplifier there is a quite old Philips TDA8510 which is out of production for quite some time so I couldn't get it from the above mentioned reliable sources. I found a quite unknown German supplier who claimed to still have these parts in stock but now I assume they also bought counterfeits from some dubious source. None of the ICs worked and I nearly destroyed my speakers due to the various malfunctions. When checking these ICs in detail I found slight hints in the case markings that they were counterfeits.
As this IC came in various configurations with inverted and non inverted channels after some research I finally found one chip variant still available where it was easy enough to modify the wiring of the PCB. I bought one from my usual sources and tell you what - it just worked like a charm.
So better throw these non-working 34063 (counterfeits?) away and buy them from a reliable source. It's a part which costs less than 0,40€ even as a single piece for end users at Reichelt...
The 680pF is physically larger: I had a close look at those photos as well, and could not see anything that seemed out of place. Thanks Torsten for the explanation.
At this point, I think my money is on the 34063 or the 680pF cap - this is because the MOSFET is not switched when the 34063 is out of circuit.
I am discounting the inductor at the moment: Even if the inductance was below spec, it would provide some boost, so at the moment it appears that there is no switching at all going on on the 34063: The main causes for this could be the 34063 itself, or the timing capacitor, which is the 680pF.
I am not aware of the 34063 being highly ESD sensitive, but they do seem to be giving problems at the moment. I of course make these clocks for myself and have not noticed that they have been faulty, but I'm starting to test all of the main components because there have been a few "hard to solve" problems with the (really totally standard) boost circuit.
Luckily at least everything is socketed, so testing is not a hard job.