Feedback as requested:
Well, first of all, the wording is generic and has some redundant text.
If I were to read this tender, I would assume whoever published it is
either technically incompetent, or has been told by a third party to
publish the tender, and would be ill-equipped to offer guidance to
whoever wants to address the issue, including regarding the causes and
potential avenues for a fix. That's off-putting.
Among other things, the tender itself should list key relevant
"technology standards", related third-party libraries/tools,
subtrees/directories within the LO source tree etc. More specifically, I
should be seeing some paragraphs regarding pdfimport, i18util, string
mirroring, character classification etc.
I would also expect to see text explaining the significance and public
'visibility' of this issue: This single issue prevents LO from being
used to edit, or even view, PDF files with content in any of the RTL
languages (Persian, Arabic, Urdu, Hebrew etc). It is a big deal.
Also, perhaps it should be made clear we are expecting a robust fix, not
some kind of reverse-the-reversed-text hack, or using single-character
text runs or some such silliness.