This is a megapost. The TLDR version is: knowledge is conditional. Cosmological observations are solid, cosmological models are solid (in a different sense), but there's always room for surprises betwixt the two. As it stands, our models are well tested and you can see some of that evidence for yourself.
There is a point here about the way science operates.
There are inarguable facts from experiment or observation: distant galaxies display redshifts and the whole sky glows faintly in microwaves. There's loads of other predictions, but those are the biggies.
Separate from that, there's the models that scientists use to integrate these observations. These are mathematical models, and contain loads of things that are not directly observed. A trivial example: my own direct observations right now are about my retina's light sensors and the touch sensors in my fingers as I type. I have a mental model of an external world that includes a laptop, and that neatly explains all of those stimuli - but strictly, that's an inference. I might be trapped in the Matrix, for example.
Scientific models are the same principle as these mental models. We have a lot of experimental results that we try to fit into a coherent worldview. It turns out that maths is a really powerful tool for that, lacking the imprecision and ambiguity of natural language. And the simplest model that explains cosmological observations is the Friedmann equations, which predict an infinite universe (or a closed finite one, but that's pretty strongly ruled out by details I'm not going into here). Friedmann actually wrote down the equations in 1922 (from memory), and cosmological redshifts weren't seen until 1929, and the glow of the sky (the Cosmic Microwave Background) not until the 1960s (accidentally, by a couple of engineers, much to the annoyance of scientists who were trying to scrape together grant money to look for it and got scooped...). So those are cases of observations being made that tested the model, and it passed those tests.
So there are three things going on here. One is the experimental data. That's a fact of life - deal with it. The second is the model. It's perfectly coherent mathematically, and can reasonably be described as containing an infinite space in which distances expand. If that is problematic for you, can I suggest that you probably don't understand mathematical concepts like infinity as well as you think you do. They are subtle and can be extremely counterintuitive - for example, I've already shown in this thread that the set of integers is the same "size" as the set of even numbers, despite integers including even and odd numbers.
But the third thing is that it's reasonable to ask whether the model is a completely accurate description of the universe. It's possible it's not. However (and it's a very big however), it's a model that is completely consistent with everything we know, and is as simple as we can make it. If you don't like the infinities then you need to add things to the model. To posit an edge you need to say that the physical laws are different elsewhere, even though they don't appear to be different anywhere we can see. Or you need to add some effect that somehow cumulates to produce an edge out of our sight. And you need to do this in such a way that it doesn't break the predictions of the model where it's been tested.
I'm not actually expecting @AStudyinPink to do that - I'm just pointing out the implications of her position if one were to take it as a serious proposition. Building new theories and models that are consistent with the limits imposed by experiment and observation but have features that a person might consider desirable is hard. So at the moment (with a few question marks), the simplest interpretation of the evidence is the standard model of cosmology (Friedmann's model of spacetime plus more detail in how particles interact, sometimes called the Lambda-CDM model) points to an infinite universe.
I would say that rejecting the opinion of experts because they are unable to fully describe their area of expertise unless you become an expert yourself is rather silly. If a topic could be explained completely in simple terms then you wouldn't need to be an expert in it to even try to explain it. That's true in any field. But, it's reasonable to ask: should I believe this person who claims to be an expert? I'll offer an opinion on how to answer that. I'd say that you should look at whether they are forthcoming about the limitations of what they are telling you. And I would say that you should look at their track record of predictions.
Having said that, I'll observe that I've just spent waaaay too long trying to make a distinction between observations (solid - they are what they are), the maths of the model (self consistent), and the comparison of a model to reality (we haven't seen any differences, so we accept it the way I accept that there's a laptop in front of me). And I'll add that there are open questions around the nature of dark matter and dark energy, and the very early universe that may one day change our models. I'll also observe that I deliberately mentioned a couple of predictions this particular model made that you could look up. You could also read about the CMB fluctuations and their power spectrum (www.astro.ucla.edu/~wright/CMB-DT.html) and see how well the Lambda-CDM model line matches the data points, and look up Big Bang nucleosynthesis (en.wikipedia.org/wiki/Big_Bang_nucleosynthesis) which shows good agreement between theory and experiment, with some question marks around Lithium.
So (at least by my own measure 
) I'd say that there is plenty of evidence that cosmologists know what they're talking about, whether it makes sense to a lay audience or not. Of course all knowledge is conditional on new information coming to light. Morpheus might pop in here with a red (blue? I forget which one took you out of the Matrix) pill, and we might find something to improve on our cosmological models. But until that happens, "the universe is infinite, and space is expanding" is the best I can do without maths.