The "YEPPICK" problem arose from a build of a snake game I released in the early 2000s whose CHR ROM had an alphabet in insular script, commonly associated with the Irish language but also used for Anglo-Saxon (Old English). One form of the letter 'r' in insular script is 'ꞃ', which somewhat resembles the roundhand cursive form of 'p'. This thread on a fountain pen forum discusses open 'p'.
I even found one scholarly reference that nails the "YEPPICK" problem. From Palaeographia Latina, edited by Wallace Martin Lindsay, page 39:
Statistics on the use of "open" 'p' have not yet been collected (except that in Beneventan script only close 'p' is reported), but it appears that a very open form of 'p' is a mark of antiquity, as in "Introd" facs. 142 (from the Anglo-saxon minuscule of the Moore Bede, written c. 773, with 'p' often dangerously like Insular 'r'.That's why when I made another insular pixel font three years ago, I chose the less confusing forms for 'r' (ʀ not ꞃ) and 's' (s not ꞅ), both of which are attested in insular script, but retained the distinctive insular forms for "d f g t" (ꝺ ꝼ ᵹ ꞇ). I don't have the old snake game handy, but here's the 2014 insular font: Incidentally, the name Mackenzie derives from a similar misreading, in this case reading yogh (ȝ), a letter derived from a form of 'g', as a tailed 'z' (ʒ). In Middle English and Middle Scots, it represented several sounds now spelled as 'g', 'y', 'gh', or 'ch' as in "Bach", especially the 'ugh' sounds in "dough" and "laugh" that have mutated or dropped out in modern English. "Coinneach" is the Scottish Gaelic form of the name Ken, where "MacCoinnich" means "son of Ken". It mutated to "Makenȝie" in Scots, which was supposed to be pronounced "Makenyie" but ended up misread as "Mackenzie".
There's a "TERRICK" floating around as well, presumably from a misreading of a capital Y with a shallow valley as a T.