I understand what you are saying, but surely the design would have to be very marginal before this had an impact on the following section of the power supply?
My point, as I said, is that
You have to be very careful when making such very general statements. That is all. In most cases you can get away with it but you have to be aware of these issues because they can bite you.
Electrolytics degrade over time. Sometimes not much, sometimes a lot. About 45 years ago the company I worked for imported two machines from America. I went there to do the training etc and the machines were shipped here and installed in customers premises. Almost immediately one started giving problems and it turned out to be a faulty capacitor which was marginal in the USA but insufficient here. I replaced it not only with a good one of the same value but with a higher value. It was a linear PSU and headroom was tight. With 60 Hz it still worked, with 50 Hz it was starting to fail, with 400 Hz it could have gone on for years.
With tolerances, failures, etc. the question is whether you would say absolutely and with no qualifications to a medical life support machine manufacturer or to a Boeing assembly plant "yeah, sure, plug it in as it is, no problem".
Those machines needed some weird adaptations. Instead of changing the electric motors (120 v 60 Hz) we installed a transformer to lower the voltage from 220 V 50 Hz to 100 V 50 Hz because at 120 V they would overheat. And they turned slower so we had to change the pulley diameter and make it bigger. The entire adaptation was quite complex and a bit on the Mickey Mouse side of things. The minicomputer read the program from a paper tape which took forever. Good times.
The loss of headroom with diminishing frequency is something which most of times will not cause problems but the engineer must be aware if life or limb of Boeing shares are on the line.
In summary, you should be very careful when making general absolute statements. And doubly so with electrolytics.
To quote myself:
You have to be very careful when making such very general statements. A SMPS designed for 60 Hz will have greater ripple at the capacitor when connected to 50 Hz. Most of the time the capacitor will have enough margin but if the capacitor is very tight initially or if it loses some of its capacitance it could lead to problems.