Thanks for this, Woelen - interesting points you raise, quite a few of them new to me. The evidence certainly suggests that Fe (VIII) does not exist. However, I remember reading (quite some time ago) why KBrO4 and the perbromates could not possibly exist - but that didn't stop them from existing. I also noted a detailed description of OsF8 - and subsequently a report saying it definitely does not exist.
I am curious then as to what might be happening when Fe2O3, KOH and a large excess of KNO3 are heated. (Unfortunately, I didn't copy the equation that Mellor gave - but this was many years ago) I remember clearly that a dark green melt was produced, that it turned white when it cooled and solidified (contrary to what Mellor said). I cannot think of any iron compound which would behave in such a way under those conditions. There are a number of iron compounds which are white, but I can't think of any which would fit this particular bill.
Unfortunately, I'm not in a position to try this experiment again.