Pure raw fertilizers will also under report a bit - because they need to ensure all batches are above the min guaranteed analysis - but this will usually be for a significantly smaller value and will be fairly similar among all products. You can see this in the Oregon database where pure raw salts - which are also registered - will have really small errors Vs lab analyses.
As if the situation wasn’t complicated enough for the average clueless grower like me flailing around in the dark trying to get my plants to grow nicely ...
I noticed an additional issue in your HydroBuddy videos. When you're adding liquid fertilizers into the HB database, you're adding them using the label compositions (which are %w/v) as if they were normal salts (which are % w/w). Although HB does have a checkbox to allow you to specify whether something is a liquid or a solid, note that the % composition is not converted to %w/w from %w/v. The software is vague in this respect - it is never specified what the % calculation type is - so I will add a note for this in future versions.
What this means in practice is that whenever you put a value into HydroBuddy for a liquid, you should divide that by the density so that you can have the correct composition. This can make a big difference. For example in the case of the GH Flora Bloom:
P as P2O5 is 5% w/v, density is 1.162g/mL
P as P2O5 is 4.31% w/w
This means that you're probably over-estimating composition by around 10-20% depending on the density of each fertilizer you use. Percentages, as almost all dimensionless units, can be pretty tricky to interpret
