Should scientific papers with no clear factual errors or associated research misconduct be editorially retracted based solely on irreproducibility?

There have been a number of high-profile cases of retraction involving papers reporting important breakthroughs subsequently found to be irreproducible. In cases such as that of the ‘Arsenic-life’ bacteria, the papers have no identified major factual errors or associated scientific misconduct, and it is debatable whether editorial retraction is justified. I contend that retractions, based solely on irreproducibility, would be appropriate in principle. This is because interpretive errors have occurred at the very least, and these errors owe largely to inadequate or erroneous data/findings generated by the authors themselves. The literature should thus be corrected to maintain validity and accuracy.