Three quantities still rest on sources nobody can read

[realmarcin]

Measured for #7 items 2 and 5, which ask a human to confirm the physical-property
values and to check whether the ~20,000 flagellin figure is actually in Macnab
2003.

It is not, and neither are the others. All four cited sources are
abstract-only, and none of the numbers appears in the text that is reachable:

record	claim	source	readable as	number present?
flagellum	filament DIAMETER ~20 nm	DOI:10.1146/annurev.micro.57.030502.090832	abstract	no
flagellum	filament LENGTH 5–20 µm	same	abstract	no
flagellum	ROTATION_SPEED ~100–300 Hz	DOI:10.1146/annurev.biochem.72.121801.161737	abstract	no
flagellum	stoichiometry "~20,000 subunits per filament; 11 protofilaments"	same Macnab review	abstract	no
ribosome	70S particle DIAMETER ~20 nm	DOI:10.1128/MMBR.00013-07	abstract	no
ribosome	MOLECULAR_MASS ~2500 kDa	DOI:10.1038/nature08403	abstract	no

#7 already flagged the ribosome mass as "the weakest"; that judgement is
confirmed, and it extends to five more.

SEDIMENTATION_COEFFICIENT: 70 is not in the table — it is definitional (the
particle is named for it) rather than a measurement needing support.

What this is and is not

The numbers are not wrong. They are textbook values and most are probably in
those reviews' full text. The defect is that nobody in this project has read
the sentence that supports any of them, which is exactly why #91 removed four
properties of this kind from the injectisome and divisome. Those two records now
carry seven properties with verbatim quotations (#65); these two carry six
without.

Options

1. Re-source against readable literature, as physical_properties are absent on the injectisome and divisome records #65 did — find open full text
   that states each figure and attach a verbatim snippet. Preferred.
2. Obtain the reviews' full text (both are Annual Reviews; neither is in PMC).
3. Remove them, as Numeric physical properties were asserted from reviews I did not open #91 did, and re-add when sourced.

Not doing any of these unilaterally: these are PROPOSED records and #7 is the
human-review track. This issue is the inventory that decision needs.

just evidence-audit will show the readability; just evidence-verify covers
any snippet added.

[realmarcin]

Retitled: six → three. PR #172 re-sourced four of them against open-access
full text, each with a verbatim quotation.

claim	status
flagellum stoichiometry ~20,000 subunits	✅ DOI:10.3390/biom11020186
flagellum filament LENGTH (upper bound)	✅ same source
flagellum 11 protofilaments	✅ DOI:10.1038/s41598-019-51440-1
flagellum ROTATION_SPEED	✅ DOI:10.1128/jb.00353-25
flagellum filament DIAMETER ~20 nm	❌ still unsourced
ribosome 70S DIAMETER ~20 nm	❌ still unsourced
ribosome MOLECULAR_MASS ~2500 kDa	❌ still unsourced

Reading the sources produced two corrections that would not have surfaced any
other way:

• The rotation-speed context attributed the whole 100–300 Hz range to light
  load. The source separates them: ~125 Hz is the high-load figure with the
  filament attached, 300 Hz or faster is low load.
• The 11-protofilament claim is not universal — Campylobacter jejuni has
  seven, and the snippet now carries that exception.

The three remaining were searched for and not found in open access. They stay
here rather than being removed unilaterally: these are PROPOSED records on the
#7 human-review track, and the numbers are almost certainly correct — what is
missing is anyone in this project having read the sentence that supports them.