Although there is immeasurable knowledge, the great scientists have always insisted on the need to quantify the phenomena to understand them. What do you think?
Most of our measurements are (or are derived from) conventions. As long as said conventions provide affordances, go at it and have a blast. But if we get trapped in our measurements, getting reluctant to switch them up for more convenient appartuses, then they are not only worthless but also dangerous. We should keep in mind that measurements are representational not ontological (or ontic).
Our measurement systems are based on a reference and on arithmetic divisions of this unit, it is the metric-decimal system. If a phenomenon is not based on arithmetic divisions, but on other logical divisions of the unit, then the measurements we make will not even be representative. If we take many measurements and treat them statistically, the results will be erroneous. If the phenomenon is not purely Euclidean we will be complicating the problem. I believe that in order to measure correctly you have to know exactly the rule.
You have to state the 'rule', yes (that means both know it yourself and be able to communicate it to others). So your measurements can be replicated. But that's all. There is no discovery of the rule (ie the rule has no ontology). No intrinsic correctness or erroneousness of a measurement on the phenomenon.
Knowledge of a rule and accurate measurements based on it might tell you nothing about the phenomenon (uninformative); or they might tell you something, but you can't employ the knowledge to manipulate the phenomenon (inappropriate informative); or they might tell you something and you can employ the knowledge (appropriate informative). In all three cases measurements would have been accurate. But that's not enough.
So what's crucial is informativeness for start. And then appropriatness on top of that. There might be multiple arguments over what metric is more informative or more appropriate and those conversations are fruitful. On the other hand, if there are many arguments about measurement accuracy, well that kind of makes me suspicious over whether metric/rule itself is informative. If people can't agree on what the rule that produced the measurements was, they aren't actually 'informed' are they?
If people can not agree on the rule that produced the measurements, then they know very little. They can have many measurements that will be uninformatives; or inappropriate informatives.
When a dog jumps up and catches a ball, its successsful action can be described in terms of a solution to differential equations involving the upward angle and velocity at which the ball was thrown. Smart dog, doing lots of sophisticated measurement and calculation? Nope, the dog just does it because it's built that way. Who's to say that some alien species in a galaxy far, far away, might not have evolved to do science in the same way that our dog catches a ball? 👽
Maybe on another planet science is done by instinct, but here we have to take measurements to do science (human). We need a ruler and count. Sometimes we can make mistakes when counting, but we will make more mistakes if we do not know exactly what the measuring rule is.