Disclaimers: this analysis is done by somebody with Bachelor’s in Mechanical Engineering and not by a structural engineer with experience in car suspension design and dynamics (but might be because of that even more damning), so take it with a grain of salt: I have been wrong before!
Let’s see a picture of real Model X rear suspension after a failure of some links.
I marked up the two broken links. These links can only be loaded in tension and compression – pinned ends. It seems that both have failed, most likely one was first then the other followed. I do not know much about the dynamics present in vehicular suspension, so I follow basic knowledge on linkage and mechanisms. Both failed links are not the ones carrying the most load in this mechanism. I looked at the upper longer link and found this to be designed in a very inefficient WTF way.
The link in question can only be loaded in tension or compression. Tension is much more straight forward then compression as buckling can occur. Let’s see my recreation of that link in 3D modeler software Fusion 360 by AutoCad. If you do not follow basics even fancy 3D Cad software can not help.
This isn’t a copy of the original but just model of its design features approximating the real thing. I do not know whether I am correct but what I can make out from the picture seems to recreated here. The link is extruded with its profile then “sliced” to make a link. If the designer added these v-shaped ‘braces” to the profile to add strength or prevent buckling we don’t know. these can be reasons, the other being that the extrusion process required it but not necessarily this way.
I did run FEA analysis for stress and deformations with very coarse mesh (large size of an element). The smaller the more accurate results. In all subsequent analyses, all parameters were the same and only geometry changed. 
This is the link in tension. yellow parts are in greatest stress. The red tag denotes the place of greatest stress. Notice that the v-shaped braces are not carrying the load. The factor of safety is 1.119 (how close the link is to failure, bigger the better)
The same link in compression. The factor of safety is 1.015. Again the v-shaped braces do nothing. I have a tendency to drill down to the basics and experiment with alternatives.
This is my alternative to the Tesla link. The factor of safety went up to 1.838 (164 percent of Tesla’s design).
Here, we are in compression. The factor of safety is 1.838 (164% of Tesla design)
Let’s see how susceptible these two designs are to buckling under compression.
Tesla’s Link
deformation at 11.85 Load
and my simplified design
deformation at 5.58 Load.
The simplified design is 147% more susceptible to buckling, the only problem is that it would have failed in either way (tension, compression) before it would be deformed to buckle at all. The factor of safety of one means you are failing right there.
This is proverbial WTF! Had the designer wanted to just guard against buckling he wouldn’t have created the convoluted design violating the basic law of constant strength throughout the part. The buckling analysis gives a good picture of how the part deformes differently on top and bottom.
Properly designed link (a bit more expensive but not much) with the thin web connecting the top and bottom. 21 Loads at buckling.
The design of the link might be a proof that indeed INTERNS designed $100-140k car because neither cost or weight or guarding against buckling or ease of manufacturing can justify making ludicrous design mistake like this.
That’s all folks. 4:30 AM. This is basic engineering shit! Believe me.
Designed by idiots.
Manufactured by shonky company.
Quality checked by nodding duck.