If the scientific method evolved what more could it become?
A better account of what scientists actually do when they are doing science?
I'm not sure what you mean there. Confronted with a problem that science can't solve?
and based it’s last resort not on observation, but to show how an experiment could lead to an observation.
That latter sounds like a question from the philosophy of science. There's quite a bit of interest in the epistemology of experiment and in what observations actually are. (Just think quantum mechanics.)
an experiment with no hypothesis can exist if we look only to observe the reaction.
Yes, I think that quite a bit of scientific work isn't really a matter of trying to confirm hypotheses. (What does
confirm mean?) There are all kinds of initial surveys for example, in geology, astronomy or ecology, that are just trying to describe what's out there, without trying to confirm any hypotheses about it.
without an hypothesis we rely on pure observation.
Yes, I agree.
But there will be preconceptions that go into it, about what will likely be observed and how best to observe whatever it is. And scientific observations often require instruments and methods, all of which have lots of theory already baked into them. So it already starts to form into loops and circles of reasoning right out of the gate. The so-called "theory-ladenness of perception". It's a philosophical question to determine whether a pure observation is even possible. Perhaps the closest that we can come there is everyday "anecdotal" experience.
An ecologist might want to observe the numbers and diversity of insects living in a particular area, in a rainforest floor or something. He or she will already have the idea of observing insects and have some means to identify them however roughly. That will depend crucially on what is already known about insects and their taxonomy.
But our ecologist might not have any hypotheses about what he of she will find, beyond the expectation that insects will be present. The purpose of the survey is to determine how many insects are present and what kinds of insects they are. That work might then enable hypothesis generation about what kind of insect ecology is happening there, what the insects are eating, what is eating them. and what sort of lives they lead.
that is the way science should be done from now on pure observation.
I wouldn't go that far. Surveys are just an initial step. They are indeed science despite the fact that they don't conform to the "scientific method", but I don't think that it would be advantageous to eliminate hypotheses entirely.
For example, a Mars rover like Perseverance can roll around taking rock samples and making preliminary analyses of what kind of rocks they are. Igneous? What kind of minerals are observed? Sedimentary? All kinds of stuff like that, without any preexisting hypotheses about what it will find. Observation in your sense. Of course all kinds of assumptions will be built in such as the assumption that the rover will encounter rocks, rocks that are close enough to Earth rocks as to be identifiable by standard methods.
But hypotheses are bound to start popping out of planetary scientists' heads about what kind of geological (is that the right word on another planet?) processes gave rise to what the rover found. So subsequent rovers (and eventually humans) will be tasked with trying to confirm those hypotheses so as to generate a better picture of Mars' history.
Observing a problem and experimenting a solution.
That sounds like the standard orthodox 'scientific method', where the proposed 'solution' is the hypothesis.
I personally think that there's a preliminary stage that is nevertheless undeniably part of science, in which problems originally arise.
Observe (the survey stage), ask questions about what is observed (a problem generation stage),
then start generating hypotheses and start thinking of ways to differentiate between them.