Accepted answer
Specifically, a certificate without a method section is giving you a claim rather than a measurement, and claims are not evidence.
The certificate is a snapshot and nothing else — it tells you about the material on one day from one sample of one lot, and extrapolation beyond that requires logic.
To be exact about it, the field you have least reason to trust is the one that looks most impressive, because appearance and accuracy are not correlated.
Published inter-laboratory comparisons of the same material routinely find spreads of half a per cent to a full per cent, which sets realistic expectations.
The caveat is that a certificate is not a guarantee, it is a data point, and treating it as anything else is asking for disappointment.
The practical summary: check the lot number, ask for the method, and reserve judgement until you have both.
edited 14 Jul 2024 by Dr_Sara_Kuusela — removed a claim I could not source