Is the singularity a physical region, or the point at which the theory stops?
What was proved in 1965 is that worldlines end, not that infinite density exists anywhere. Which of those the corpus is describing is an interpretation, and the cosmic censorship conjecture — that such regions are always hidden behind a horizon — has never been proved. The Journey features the singularity as a limit on general relativity rather than as a described object, and does not choose between the readings.
Does anything happen at the horizon?
Classical general relativity says nothing local marks the crossing, and that is exactly the prediction some proposed resolutions of the information problem are prepared to give up: the 2012 firewall argument concluded that an infalling observer might instead meet high-energy quanta. No observation distinguishes the possibilities and none is currently conceivable. The corpus raises the firewall in two uncertainty fields without a citation, deliberately, and this package carries that gap rather than filling it from anywhere else.
How much of the published image is in the data?
Eight stations sample only a sparse scatter of spatial frequencies, so most of what a filled aperture would gather is missing and the published image is one reconstruction among many consistent with the data, chosen under stated assumptions. The strongest answer available is not an argument but a repeat: the 2018 epoch gave the same ring diameter with the brightness maximum in a different place. For Sagittarius A* the concern is sharper, because the source rearranged itself during the observation and independent reanalyses have questioned specific ring features.
Did the objects that merged in 2015 have horizons?
Nothing in a gravitational-wave signal establishes it. What the data show is consistency with the Kerr prediction. The ringdown of a single event is short and faint, so how tightly current observations constrain the no-hair result is contested, and claims of more than one quasinormal mode in individual events have been disputed.
What powers the M87 jet, and how well is any spin actually measured?
Whether the jet draws mainly on the rotation of the hole or on the accretion flow is not settled observationally, and measured spins remain model-dependent. The thin-disc description written in 1973 does not apply to the flows the array actually imaged, which are hot, geometrically thick and radiating inefficiently far below the Eddington rate — so the most familiar model in the subject is not the one the famous images show.
How did S2 get where it is?
A massive young star should not easily form that close to a supermassive black hole, and the corpus records its presence as unresolved. This is more than a curiosity for this Journey: the same star supplies half of the strongest cross-check in the subject, and the object doing the checking is itself not understood.
Where does the information go?
Open, and the last chapter exists to keep it that way. Page made the problem quantitative in 1993 by specifying the entropy curve any unitary account has to reproduce; since 2019, calculations including previously neglected saddle points of the gravitational path integral do reproduce it — without identifying the mechanism by which anything leaves, and with the people who derived them disagreeing about what has been shown. Hawking conceded his own bet in 2004 on grounds that were not generally accepted, and Kip Thorne never conceded at all.