Actually theories are abstractions of the relationships between concepts that are only indirectly-measurable, while hypothesis are the more concrete understanding of the world derived from empirical evidence and link that which is measurable to that which is not. Without some level of indirect-measurement required there is no need for a theory, we would simply have fact... such as the fact of microbial evolution, the fact that DNA exists and so forth.
This means that theory is not something verifiable through observation, but a systematic method of understanding complex reality in a way that is parsimoniously comprehensible. If competing theories have also yet to be disproved then there is no 'right' theory, only a trade off between utility and falsifiability -> the more general the theory, the less well it is defined concretely through variables and hypothesis -> the better it is at abstracting reality and the worse it is at being falsifiable.
The point being that a "theory" is neither the super-hypothesis that you seem to think it is, nor is it the half-witted conjecture that the gpp thinks it is.
It is a multidimensional abstraction of reality that, while useful for explanation and at some point empirically disprovable, must reside at some level of abstraction and thus make trade offs between its usefulness and dis-provability.