Cannot curl the page pdf action endpoint to directly download the pdf

Summary

Up until Confluence version 1000.150.0 exporting a page to PDF by clicking the 3 dots menu and selecting "Export to PDF" option would create the PDF and present the option to download it within the same page (without refreshing the page or loading another page).
However, from Confluence version 1000.161.0 forward the export page to PDF is considered to be a long running task, thus when a page export to PDF is selected it loads the "long running tasks" page before presenting the download link.

Environment

Confluence Cloud

Steps to Reproduce

1. Go to a page click the 3 dots menu
2. Select the "Export to PDF" option

Expected Results

A direct download of the PDF.

Actual Results

The long running tasks page is loaded and only then the export is processed and it produces a link instead of the direct download file.

Notes

This breaks Bob Swift's CLI, specifically the example 123 on this examples pages.
Also, if you used cURL to automate PDF exports download this no longer works as cURL returns the HTML of the long running tasks page.

Workaround

Click the Download here link in the Long Running Tasks screen.

Or if you use automation you can make use of the longtask method of Confluence's REST API to get the link to download the file.