We can think of many things as having types and tokens, for example we can easily separate the idea of a 'number' and all the many quantities that use that number, why do we struggle to separate the idea of a type from a token? For example if I ask someone to tell me how many digits are in the numeral '111' they may say '1' or '3', The word letter is also ambiguous in this way, We have the letter 'A' and the different letter 'B' but yet 'Boot' is still called a 'four letter word'. Why do we not simply define the 'letter' to be the unique symbols 'A', 'B' etc, and refer to each token simply as a 'token' or 'occurrence' of the letter, similarly for digits, it seems strange that Logicians and linguists have not simply make this distinction clear.
"Why do we struggle to separate the idea of a type from a token?"
Well, in ordinary language, there is no such struggle. It is typically clear via context exactly what someone means, and if not, clarification is easily offered, and even if not, there are no weighty matters (ontological import!) hinging on the distinction.
To my knowledge, there is no deep struggle in logic about this distinction, since in fact it is usually perfectly clear what a logician means by letter, typically that very mark on the page. For linguistics the distinction is more important. There are at least two reasons why linguists do not define such notions. First, linguists have not come to agreement on what a type is. In particular, while most linguists (implicitly perhaps) accept phonetic, morphemic, and lexicographic types, it is unclear whether there are semantic types. Secondly it is unclear whether is it their job to- a linguist is typically paid to test empirical hypotheses about language. For more, see https://plato.stanford.edu/entries/types-tokens/#Lin. But perhaps they should be, see Hutton 1990 (referenced in source above).
Further, tokens are not occurences, for that argument, see Can something be both a type and a token?