@jaror You're incorrect. There's at least three major distinct preferred chronotypes among humans; the diurnal one has gained legal supremacy in parts of the Western world mostly for religious and religion-adjacent reasons. Diagnosable DSPS does have known adverse health factors associated with it, but there's good reasons to suspect that these are not inherent, but an outcome of having a sleep schedule different from the dominant leglly preferred one; effectively, a form of minority stress. If a nocturnal chronotype was legally dominant, minority stress would be likely to ghe other way. Perhaps unfortunately, doing this on a society-wide basis is uncommon;, with the only major 20th century example being Stalinist Russia, but, well, even if we could study that one's biochemical outcomes, there would have been a lot of hard-to-etricate confounding factors. The arguably next context closest to imposed nocturnal activity lifestyle involves shift work, mostly factory work, but because of legal preferences for diurnal activity are generally dominant, all major jurisdictions require shift most workers to be regularly switched between shifts, which seems to have adverse health factors of its own. It can also have a diagnosable name, the Shift Work Sleep Disorder.
All in all, the limited current knowledge suggests that, if the political decisions were different, a situation with the nocturnal chronotype being the legally preferred norm and people with the diurnal preference being diagnosable as Premature Sleep
Phase Disorder sufferers, with the asociated health rigsks largely similar to those currently associated with DSPS, with only the SAD risk — which is associated with sustained lack of sunlight exposure during one's waking hours — being significantly diffrent, would be entirely consistnt with human physiology.

Preference for a form of biphasic chronotype has legal protection in Spain, largely as an alternative to having to invent office air conditioners, and it's less studied than the other two ones. From wha we know, it is likely ha most people wih a biphasic preference have easier time adapting to either diurnal or nocturnal activity phase, and vce versa, so it is likely that these differences between one's internal preference and societal imposition lead to overall lower level of minority stress than the differences between living diurnally in a nocturnal-dominant society. Even the diagnostic criteria for a biphasic sleep to qualify as a 'syndrome' are harsher than those for a nocturnal chronotype to qualify as diagnosable DSPS. But really, we don't know enough about biphasic activity chronotypes, and More Study Is Needed(tm). It may be that genuine and stable biphasic prfrence is less common than either diurnal or nocturnal prefrence, but in light of the evidence for higher capability of mask, even the real prevalence of biphasic activity preference is not really known yet.

@jaror As for your first question, basically some simple-minded start-up-bros noticed around ca. 2010 that occasional all-nighter experience was kind of common among effective hackers, and, like the islanders enjoying all the fascinating cargo that their new friend John Frum brought in during the WWII, they erroneously postulated a causal link, and envisioned that if they made all their employees do regular all-nighters, everybody would be a particularly effective programmer. It really doesn't work that way, though; and unlike John Frum's friends saddened by the end of the war who had to build their own flight strips and their own control booths and their own radio antennae, from bamboo, who merely not get the cargo they're trying to attract, forcing people on regular diurnal shifts to also do all-nighters on a schedule bears worse health outcomes than either stable diurnal or stable nocturnal activity phase on their own.

But, on one hand, it's hard to measure a programmer's work product and easy to fool oneself to think that the last arbitrary change made things better, and on the scond hand, the start-up-bros tend to not see the health problems they cause to their employees as their problems, so the stupid practice has found wide followership among start-up-bros and, more recently, among some middle managers, and will probably not be abandoned until a major lawsuit finds it illegal enough to merit meaningful financial damages from a company practicing it.