Mixing the loudest, densest section first sets the headroom ceiling the rest respects
The climactic section carries the most elements and the highest energy, so mix it first to a competitive standard under worst-case conditions (all instruments playing simultaneously). This sets the mix-bus ceiling, dynamic peak, and master-buss behavior; every other section is then refined within that established framework. If you instead start from a quiet verse, the levels and processing chosen there will likely need significant readjustment when the louder section arrives — often causing overloaded processing, lost headroom, or a global re-level of everything. After the climax is finished and saved, build the remaining sections by muting all tracks, looping a new section, and reintroducing tracks in rank order, using mults or automation wherever a track must change in the new context. Mixing the most energetic moment best also has the greatest impact on how the whole song is perceived.
Examples
Get the final chorus to commercial standard first, with master-buss compression working at an appropriate level; then mix the sparser verse to sit perceptibly lower without resetting the ceiling, leaving the chorus as the dynamic peak.
Assessment
Explain why mixing the loudest section first produces better results than starting from the quiet intro: what headroom problem does it prevent, and how are the other sections set relative to it?