Bad example on your part. We can directly observe the Sun, and as you note, directly measure neutrinos. Still, solar fusion is just a theory, just one with no controverting data.
However, the planetary accretion processes cannot be directly observed. Yet the models do have controverting data, which I cited, in the form of reproducible calculations.
It's simply unreasonable to ever say we "know" a theory to be true when someone can demonstrate the impossibility or improbability of the theory, as has been done with all planetary evolution models to date.
"Scientists know' can be shorthand for 'the established scientific consensus allows us have a very high degree of confidence."
See, you're doing it right now! You don't want to say "We don't know." It sticks in your craw. Are you a scientist? ;)
Other euphemisms scientists often use for "We don't know":
"It isn't clear..."
"The best evidence indicates..."
"The consensus is..."
One thing no true scientist can forget: science is not a consensus enterprise. If one million scientists hold to a theory, and one scientists -- or even a non-scientist -- can provide reproducible calculations or experiment contradicting the theory, then the theory as posited must be discarded.
Moreover, to even qualify as a scientific theory, the theory must be falsifiable. Planetary accretion theories are falsifiable, as I've cited, but many other so-called scientific theories are not. Such as anthropogenic global warming (AGW) .