andrasz
Your theory would hold some merit if the figures weren't so out. Looking at BA's fares (to avoid any ignorant comments from others like "well Norwegian should just charge more then") from LON-NYC, the bread and butter of Norwegian's long haul operation, the revenue per passenger once government taxes and airport fees have been taken out is more like $68 per passenger per sector. A far cry from $400, and using your own cargo price estimation, about a third of the $200 per 100kg, assuming the costs associated with handling cargo and passengers are similar. Of course the average revenue per passenger would increase once you include higher fare brackets and cabin classes, but these aren't the passengers that are going to be sacrificed for the carriage of cargo.
The myth of cargo being a prime revenue source may be a myth on some routes, with some airlines. But for airlines like Qatar, for example, cargo is the reason they're able to operate such a high number of frequencies to offer optimal connecting opportunities for passengers. Qatar don't (pre-Covid) fill six widebodies per day into and out of Heathrow for around 45 of 52 weeks of the year, but they do regularly shift over 100 tonnes of cargo per day in each direction. It's not a case of waiting for the last bag to be checked in and then working out how many kg of cargo they can load, the airlines will be able to model an estimated passenger load a couple of weeks in advance and artificially adjust capacity above and below the floor as required. A vast amount of long haul routes all over the world rely on cargo to be viable, it's not just a few extra dollars to the bottom line.
Going back to Norwegian and Singapore, offloading bags and/or passengers would've been far more beneficial than offloading cargo, and this has been covered a number of times. As you well know, being offloaded as a passenger is rarely a traumatic experience that results in "100 lost customers per mistreated passenger". There's usually plenty of volunteers to be offloaded with the incentive of 600 Euros in DBC, a free hotel room and a seat on the next flight. Nor is a missing bag going to cause the airline to lose 100s of passengers. All airlines have bags that don't end up on their intended flight, intentionally or otherwise. I've had Iberia misplace my bag three times on the exact same widebody flight from Madrid to LHR, including once when no bags at all made the flight - bumped off in favour of the vast quantities of fresh fruit from South/Central America.
With that being said, if your argument was that low cost Ultra long haul doesn't work, I'd tend to agree. Singapore, which is essentially ultra long haul, was always going to be a marginal route with performance issues. Another factor at play in Norwegian's case was that most of the flights departed from Singapore in the late morning, when temperatures are typically 3-4 degrees warmer than the "traditional" European departure bank around Midnight. Not only did the few extra degrees add to performance issues, but the daytime airway closures in India add significant track mileage to flights operating Westbound at this time of day. They simply didn't have the aircraft availability to operate evening departures in both directions, not helped by the well publicised RR issues. And that's before you consider the volatile airspace the route uses, whether it's Pakistani airspace closing, Iran's airspace restrictions, Russian overflight or, if all that fails, the near impossible task of getting a Saudi overflight permit at short notice. Low cost to Singapore was probably always doomed. Low cost to New York, Los Angeles and Orlando on the other hand...