To clarify ... TFT needs to be updated to test the current RDF, RDF*, SPARQL definitions. The current definitions that were migrated are out of date.

IGUANA needs the tests defined that reflect the characteristics that were specified.

Without these changes, the infrastructures are useable - but not useful.

So, it is likely that the estimate of "8" points is low.

@dcausse I have the work item to update TFT to the latest tests on my to-do list. I can try to prioritize it higher, if that is valuable to you.

It is probably 3-4 days of work.

Completing IGUANA would require more work to get basic functionality. Maybe 2-3 weeks? (top of head guess)

TFT
does not seem appropriate for the kind of tests we have to make within the scope of the graph split project, it does not provide anything to ease the creation of test sets, I created T351819 to create such tooling.
IGUANA
does seem more promising, after a couple fixes I was able to run a simple test against our public endpoint, User:AndreaWest/WDQS_Testing/Running_Iguana#How_to_Change_or_Extend_the_Tests_and_Infrastructure gives a good overview of the extracted metrics. The test output in its current format (RDF) might not be directly actionable by a data-analyst and might need some transformation to a kind of tabular data.
I used IGUANA version 3.3.3 and had to apply several fixes (I forked in https://gitlab.wikimedia.org/repos/search-platform/IGUANA), we should consider using the new metrics added by Andrea in QPSMetric (meanQueryTime, geometricMeanQueryTime, min, max and their penalized version).
I attempted to upstream my fixes but it appears that the IGUANA maintainers no longer want to maintain the 3.x branch because they're working on a v4 version which is a major refactor, I did not have time to look too deeply into it...
For the wdqs graph split needs it might OK to use our v3.3.3 fork but for blazegraph alternatives evaluation we might consider rebasing Andrea's work on the upcoming v4 branch depending on its status.