This is a performance-related question - for instances of Jira that have a large amount of project, issues and/or custom fields.
My question has specifically to do with TEXT FIELD TYPES (singly and multi-lines) and custom field context definitions. The need to segregate single-select and multi-select list types etc. is obvious enough since different projects will have different options for these "reused" field names. Instead of using global contexts that can be shared across projects, issue types etc. our team manually adds a new context (adds to an existing context) for every new use of the custom field. For Text field types this is a lot of work for no clear benefit (assuming you don't have defaults being set, which is probably a bad idea anyway).
Does anyone have any first-hand evidence or first-hand knowledge (say you work for Atlassian on Jira) that there is good reason to go through this trouble for Text field types?