Some excerpts & understanding from:
EINSTEIN'S MISTAKES: THE HUMAN FAILING OF GENIUS by Hans Ohanian.
The book does not try to diminish the man's extraordinary achievements. Physicist Ohanian shows that Einstein was human like the rest of us. Like a real scientist trying different approaches, Einstein made many errors. Part of his greatness was the willingness to make mistakes
The first derivation of m=E/c^2 is in a 2-page addendum to "The Electrodynamics of Moving Bodies." There is an error in this hastily written proof, and more mistakes in the second, third and fourth proofs of 1906-07. Only in 1911 did Max Laue produce a full proof. Einstein's fifth proof in 1914, his sixth proof in 1934 and his seventh proof in 1946 also contained errors. By this time nuclear reactions and the atomic bomb had proved to the world that E=mc^2.
EINSTEIN'S MISTAKES finds errors in 4 of the 5 seminal papers, including the photoelectric effect and the size of molecules from Brownian motion. Einstein's PhD thesis was full of mistakes. The road to General Relativity contained many errors and dead ends. In 1916 Einstein made a mistake interpreting Mach's Principle. The last decades of Einstein's life were spent in a fruitless search for Unified Field Theory.
Attempting to unite electromagnetism with gravity, he tried many approaches which all failed. The book lists dozens of mistakes stretching across an entire career.Einstein's greatest blunder was the cosmological constant. In a 1917 paper Einstein dared to imagine the entire Universe.
According to General Relativity, mass causes Space/Time to be curved. Einstein realised that enough mass would cause the Universe to be curved into a sphere of 4 dimensions. Travel in any spatial direction would be confined to the sphere. Such a sphere would collapse under its own gravity, unless it were already expanding. To support the sphere Einstein invoked the fudge factor of a cosmological constant. If Einstein had proposed an expanding Universe it would have been one of history's great predictions. No doubt he would have been ridiculed for having no supporting evidence. That's why it is called a prediction, boys!
If Einstein had waited long enough, Edwin Hubble would eventually have proven his amazing prediction. When faced with Hubble's evidence Einstein had to admit that the cosmological constant was a blunder.
Fortunately Einstein had a great patron in Max Planck. Planck was an editor of Annalen Der Physik, otherwise Einstein's great papers may not have been published in 1905. In later years Planck was first to hail Einstein as a new Copernicus. Planck's blackbody formula was an experimental result; many years passed before someone found a mathematical derivation.
Einstein had another regret: Submitting to refereed journals. Around 1936 he submitted a paper to Physical Review, the leading American journal. Despite Einstein being by then the world's most renowned scientist, his paper was returned with anonymous comments from a referee. Einstein withdrew the paper and henceforth avoided PRL and any journal with anonymous referees. This was a blessing in disguise, for the paper contained an error.
A child could figure out that R=ct and GM=tc^3. Proposing this gets the most tangled mathematical objections: the units don't add up (they do), the metric is (1, 1, 1, 1) rather than (1, -1, -1, -1),...and so on. We can see why these people, despite their expensive education, have never come up with anything original. By their own admission, such people referee papers. If they were reviewing Copernicus, they would insist that everything be expressed in terms of a fixed Earth.
History gives us many lessons for today. Even an Einstein can make mistakes. The cosmological constant was a blunder and still is. Predicting an expanding Universe or a shrinking speed of light is worth the wait for confirmation. Don't feel bad if refereed journals reject your stuff. A great physicist is prepared to make mistakes, lots of them.
The book admires Einstein's "mystical, intuitive" approach. He used mistakes as "stepping stones and shortcuts" to success. We can still learn a lot from Einstein.