This is the same exact industry that gives you paid usage limits as a unit-less percentage bar then gaslights customers every time the algorithm running that percentage bar changes or they lobotomize an existing model with increased quantization to squeeze a few more dollars out of existing hardware.
"Failing cleanly" might make their moated hype-machine look bad pre-IPO, so they certainly aren't going to do that voluntarily.
So far I have mixed impressions, but they do indeed seem noticeably weaker than comparably-sized Qwen3 / GLM4.5 models. Part of the reason may be that the oai models do appear to be much more lobotomized than their Chinese counterparts (which are surprisingly uncensored). There's research showing that "aligning" a model makes it dumber.
That is fine as long as the input / output is always in UTC... but at the end of the day you often want to communicate that timepoint to a human user (e.g. an appointment time, the time at which some event happened, etc.), which is when our stupid monkey brains expect the ascii string you are showing us to actually make sense in our specific locale (including all of the warts each of those particular timezones have, including leap second, DST, etc.)