I saw those videos: they are interesting.

You are right, the fact that the 3D printed object are anisotropic add an extra variable to the game. You could use the strongest filament, but if the layer are oriented in the wrong direction you will get a poor results.

I don't know if to avoid such issue, it could make sense for the producer to test molded specimens of filament (cylinders or bricks). In this way they will consider only the material itself. However, the inter-layer bonding properties will anyway play a role when actually printing, so there is the risk that those number won't translate into the printed object properties. Otherwise, they could leverage their experience and claim that the tests were made in optimal conditions, so that you know that that's an upper-bound.

It's strange that they provide that information for the Carbon Capture PETG and they skipped all the remaining filaments, included the standard PETG. Maybe, they just started testing and adding that info to their products.

I am new to the field, so I cannot judge, but I am happy to hear that you noticed an increase in the mech props reporting. Hopefully, that will be the standard.

Yes, it's definitely a good point of information. It looks like it was also referenced in the Prusa blog: Advanced Filament Guide.

That post is a bit old (2020), and I fear it has not been updated in a while, considering that the post writer replied only in the first year and all the following comments ended up without replies. It is a pity, considering also that the post ended with a sort of request for users to propose new filaments to test. I see that the last comment is fairly recent (2024) so I guess someone is trying to get more information to be added. Alas, it's kind of understandable: it could be expensive and time consuming to test other brands filaments for the sake of keeping the maker community happy.

In the post seems that the data collected in the table is coming from real tests made at Prusa Research, so this makes me hope that the experiment setup has been kept the same for the different tests.

Currently, there are some vendors that provide a lot of information about some of their filaments. I cannot say anything about how accurate or trustworthy they are, though. Unfortunately, as reported here by other users, it looks like the market if filled by "cunning" sellers who are not transparent regarding how they asses the properties of their filaments. Anyway, as far as the physical properties to track are concerned, one of the most extensive table I found reported:

• Density [g/cm³]
• Elongation at Break [%]
• Tensile Strength [Mpa]
• Flexural Strength [Mpa]
• Flexural Modulus [Mpa]
• Young's Modulus [Mpa],
• Izod Impact Strength [KJ/m²]
• Melting Point [°C],
• Melt Index [g/10min]
• Heat Deflection Temperature [°C]

I could think that with those you can have some idea about how good the material will be for mechanical parts.

There are other properties which are sometimes advertised as "chemical resistance" which are often left without a reference.

Apart from the data you can get from the sellers, I found a table that, even if not providing a thorough analysis of the filaments in terms of the properties listed before, it looks like an attempt to gather together data to help user chose the right filament for their application: Prusa Research - Material Table

Thanks a lot for for corroborating the issue. Indeed, I am aware that labeling could be blurry since everyone keeps the blends used for their filament secret. This is the reason why I am interested in knowing if someone collected the numbers. Being able to compare Elastic Modulus, Elongation at Break, etc. could solve the problem of having to use the labels when picking a filament (considering also the cost $$). PLA+ was just an example, where everyone says that it's "more though" than "ordinary PLA", which I could translate with "less brittle". However, without absolute numbers the information contained in such statements is laughable.

Without numbers, either I just wait to have a bit of experience with different filaments, or I will need watch all those videos with users printing hooks and running tensile test on them.