It's an actual percentage based on dividing the number of passengers by the number of seats offered. So it accounts for differences in air craft used. I cannot say that all of these numbers are 100% accurate because I believe a lot of it is based on airlines reporting, but these numbers at the very least are a good approximation of the number of passengers and associated load factors.
It should go without saying that a load factor does not give a proper indication of a route's financial performance. Boston and London are two of the world's largest financial centers (London is typically in the top 3, Boston is typically in the top 3 in North America) so you're going to have a lot of business traffic filling up the business and first class cabins. Coupled with cargo, you could very well make a profit flying a route at 40-60% load factor and I am sure there are plenty of carriers who do so.
Unless someone works for an airline, or a consultant who works with airlines, no one will ever know the load factors on specific classes. You can make some guesses on the yield based on the fares, but that would take a lot of time to aggregate the data.
BOS-LHR isn't exactly a winner for AA, DL, or UA either just on load factor alone but they stick with it.
Of daily transatlantic flights in October 2022:
AA: Only RDU-LHR has a lower load factor than BOS-LHR
DL: Only DTW-MUC and SLC-LHR have lower load factors than BOS-LHR
UA: BOS-LHR has the lowest load factor of all daily UA transatlantic flights
There's a reason DL and UA use their premium-heavy configurations on that flight.