Polydron Frameworks is an open-frame version of the marvellous Original Polydron educational toy, and the types are compatible. After succeeding in creating printable versions of the original, the next logical step was to tackle Polydron Frameworks. The only guidelines that were available were the compatibility claim, as well as pictures on the web, which show that the mechanism differs from the original.
So, after a few iterations, we have here a collection of pieces similar to the Polydron range, with some additions:

1. Equilateral n-gons, from 3 to 10 (i.e. triangle to decagon). The decagon comes out at 222mm x 216.36mm, which means that strictly speaking I am not supposed to be able to print it on my Prusa Mk4, where the Y-size is limited to 210mm. However, sometimes one can get ahead in life by cheating and lying - it turns out that you can push the Y-size to 216.5mm, misleading Prusaslicer regarding the true dimensions by setting the bed size with a Y offset of +4mm. So we use the part where the purge line is deposited too. This leaves a full 70 microns of margin to spare on each side. (See file FW10.3mf, where I have configured it thus.) If this does not work for you, you can still console yourself with a heptagon and a nonagon, which few others are privileged to own.
2. The two delightful rhombi (rhombuses?) with long-to-short axes ratios of sqrt(2) and phi respectively. Once again I refer you to this article for interesting information about these shapes.
3. More triangles: right angle (half a square), isosceles, and root 2. This introduces a new edge length, longer than the standard by sqrt(2), i.e. the diagonal of the square. These triangles respectively have one, two, and three of these longer sides.
4. Rectangle, with side lengths of unity and sqrt(2), to go with the triangles above.

The Frameworks pieces are slightly easier to assemble, especially when it comes to inserting the last panel in large polyhedra, like an icosahedron or maybe a triacontahedron, because you can access both sides of the joint, unlike the case with the solid (original) panels. And they use less material. But sometimes solid is good, too.
My personal preference is for a combination of types - they are compatible, after all.
And the most important thing is to have lots of pieces. More is better, it turns out.

In Prusaslicer I use the 0.15mm SPEED profile, with variable layer height (at the 0.50 Quality/Speed balance).  Parametric design done in FreeCad.

As always: if the fit is too tight, consider a small negative 'XY Size Compensation' or its equivalent in your slicer of choice, or conversely a positive value in case of a too loose fit. And with small I mean exactly that - adjust by a few tens of microns at a time. Remember that the effect of these changes is doubled, because the mechanism engages with another version of itself. I suggest first printing single triangles to test and adjust the slicer settings, until you are certain that everything is dialed in perfectly.