Your Einstein example actually contradicts your point about reputation.
The mistake is using hindsight to judge that paper's quality. Instead, we have to imagine looking at that SR paper as if we lived in 1905 and didn't know if the 26-year-old patent clerk was a crackpot or a genius.
Einstein sent his manuscript to an academic publisher where Max Planck was one of the editors. This is important because MP already had the reputation in physics so his recognition of Einstein's papers' value was taken seriously.[1] The journal itself also had a reputation which was reinforced by the reputation of its editors.
In 1905, Einstein's papers didn't stand on their own. Its dissemination and acceptance was helped along by Max Plank's "stamp of approval".
[1] https://en.wikipedia.org/wiki/Max_Planck#Einstein_and_the_th...
We could even argue that a purely meritocratic system of publishing will accelerate good quality research as opposed to a reputation based system.
Funding and career advance should be ortogonal to knowledge distribution.