Many years ago it was typical that HTTP paths paralleled file system paths. eg. for a URL like
.../widgets/green/round.html
the files on the server might be like
/widgets
widgets.css
widgets.js
/green
round.html
square.html
But these days quite often the HTTP paths do not correspond to file system paths, but instead serve content based on routing.
I’m fairly certain I have seen discussion here about having a file system folder that could be accessed with an HTTP request, but I don’t recall when or where.
IIRC, it requires creating a route for it.
Does sparkpost not have another way to verify site ownership, eg. entering a hash value in a meta tag that could be inserted into your customize head?