First of all, I should mention that the two options you provide are not the only understandings, so one does not have to accept any of them. For instance, in Copenhagen Interpretation, to talk about the position of a particle before the measurement is not meaningful. This is not saying that the particle has a position but we don't know what it is, no. CI says the concepts we use in macroscopic world such as having a definite position, momentum, and so on cannot be applied to the microscopic realm. Therefore, considering a quantum particle as a point-particle with definite trajectory is a wrong picture.
Secondly, there is a distinction that a quantum system is in superposition of two states and the particles are in either of two states but we do now know which one is which. The second one is called "mixed state". Let me give you an example:
Consider two boxes A and B, each contains 100 electrons. In the box A, all electrons are in the superposition spin-state of up and down along z direction. In the box B, %50 electrons are in the up spin-state and %50 electrons are in the down spin-state. In other words, the box A represents superposition and the box B represents our lack of knowledge.
Let's make an experiment, if you make a spin measurement on the electrons in two boxes along z direction. For both boxes, you will find around the spin of 50 electrons will be measured up, and the spin 50 electrons will be measured down. However, things get interesting when you make a spin-measurement along (let's say) x direction. In the box B, you will still get %50 of electrons have spin up and %50 of electrons have spin down state. However, this is not the case for the electrons in Box A. If they are in a certain state, you can measure %100 spin up or spin down along x. Therefore, there is experimentally difference between our lack of knowledge and quantum superposition.
Unfortunately, this example is not well-constructed because it requires math. For example, I should have written the exact state of the electrons in box A in order to justify the measurement results. However, the message is clear: even if quantum states represent knowledge (that's what Qbism says), they do not represent "classical knowledge".
Additionally, the experiments showed that Bell's inequality is violated, which means if particles have predetermined values, they (entangled particles) have non-local correlations. The other choice is accepting superdeterminism but I don't want to go into the details.