Where the mark is arbitrary

The curve is a logarithmic spiral and has a reason. The radii around it do not. These three put the eye and the ring onto the spiral's own φ ladder — and the third leaves nothing chosen but one seed radius.

Today — what ships

span 207.9°, radii chosen by eye. reach/discR is exactly 3.

span
1.1550π
rInner
258.13
discR
150.00
reach
450.00
arm growth
1.7433
rInner/discR
1.7208
reach/discR
3.0000

One arm = one φ

span = π exactly, so each arm grows by φ end to end and the pair spans a full turn (φ²). Eye untouched.

span
1.0000π
rInner
278.12
discR
150.00
reach
450.00
arm growth
1.6180 = φ
rInner/discR
1.8541
reach/discR
3.0000

Full φ ladder

span = π AND discR = reach/φ². All three radii read 1 : φ : φ² — nothing in the mark is chosen but the seed.

span
1.0000π
rInner
278.12
discR
171.88
reach
450.00
arm growth
1.6180 = φ
rInner/discR
1.6180 = φ
reach/discR
2.6180 = φ²

Two bugs, independent of any redesign: the docs say the arm “grows by φ every quarter turn” in three files, but the code halves the rate — measured growth is φ per half turn. And biteR is inert at every spec above: the lobe never reaches it, so the ring you see is the spiral's own tail, not the bite.