TRENCHES is a forensic operating system for Tesla and SpaceX objectives. Percentages are estimates from public artifacts — Master Plans, earnings, S-1s, FAA/FCC, NASA, trackers — not official KPIs. If something is a demo, it is labeled a demo. If a date moved, the old date is crossed out.
A merger module is game theory, not fanfic. Elon sits on both sides of the table. China is the silent killer. Combining them “might make Elon’s life a little easier.” That is the honest executive reason.
Completion answers “is it built?” Relevance answers “does this trench change whether Tesla and SpaceX should be one company?” Terafab is 8% built and 98 relevant. That is not a contradiction.
Relevance ≠ completion
A 100% done Model 3 line barely moves merger math. A 8% chip moonshot with a $55B shared bill moves it a lot. Do not rank trenches by how finished they look in a keynote.
Terafab: 40 shared + 30 bill + 20 agency + 8 execution = 98. Highest in the set.
How the score is built (0–100)
Shared necessity
0–40. Do both companies need the same physical or legal node?
Bill neither wants alone
0–30. Capex that is easier as one P&L than as a related-party invoice.
Agency conflict
0–20. Traffic that a merger would collapse (Elon on both sides).
Execution haircut
0–10. Discount if it is still slides. Terafab keeps 8 of 10 because the LOI is real even if the fab is not.
Low completion + high relevance is a moonshot. High completion + low relevance is a closed trench that no longer moves the clock.
Ranked trenches
Sorted by merger relevance. Click a row to inspect. Gold = shared node.