@NicoleCRust Thanks for the question, Nicole. I got a little carried away...
A lack of self-awareness of the communication goal is sometimes an issue. One way I see "getting excited about science" go bad is when there is a clandestine switch of the topic of discussion, hiding a goal shift from increasing scientific understanding to increasing psychological reward.
Someone starts with a show on "The Science of Canine Cognition", which explains currently known "facts" about behavior, neuroscience, history, and so on, with equal emphasis on the process by how these facts were determined, including their uncertainty. Excitement, sure, and relating things to people's everyday experiences with their beloved pets, and keeping the scientific process in the forefront. Motivation and Method gets equal screen time with Result and Discussion. It is a show about a cool area of scientific study.
But people resonate with the doggy facts more than the explanations of experiments. So the show becomes "How Dogs Think", and the description of process gets reduced to "Scientists at Prestigious University have determined..." and everything is "explained" by videos of cute dogs doing things. The essential concepts of uncertainty and empiricism get thrown out, and the narration now contains fewer passages of questions ("How might one figure out if a dog understands English? What does it really mean to understand a language in the first place?"), and more blanket statements ("Dogs know so many words! Look at this adorable dog operate a pedal board that plays words."). It is a show of claims about dogs.
And we might keep going. The show is "Your Dog's Incredible Inner Mind". Claims are selected with a strong "excitement" bias, and explanations are stories that seem to make sense and aren't obviously wrong; maybe a scientist somewhere is willing to say on camera that it might be true. Watch this dog get over the loss of their favorite plushy by "singing" along to Taylor Swift lyrics, just like you might. This other dog writes poetry (remember that pedal board?) in an innovative Romantic-Absurdist style. No questions, no process, just examples with just-so narration. It is a show of our fascination with and emotions about dogs.
And if that last show spurs someone's curiosity and makes them want to learn more about science, great, except that the show didn't actually get them excited about science. It got them excited about dogs. The process by which we come to know things isn't part of that deal.
I'm reminded of a critique from the ancient Internet that the website "I F*cking Love Science" was better named "I F*cking Love Trippy Pictures of Multicolored Fluorescent Stuff".
To me "getting excited about science" means getting excited about the applied process, not the particular list of conclusions best supported by evidence we happen to have at the moment. I see the clandestine switch from Process to Fact behind a lot of the arguments people have over what is "good" general audience science explanation.
As scientists, we don't accept blanket statements of fact, unmotivated and unjustified by evidence, from our colleagues. We should not feel comfortable making those kinds of statements to the public. Popularizing carries the additional burden that you have to explain process at the same time as you explain the subject, because your audience in general has not been trained on the relevant conceptual toolkit. To me that is the core thing that makes it hard. To avoid the failure mode I tried to exemplify above, it's critical at the outset to be clear on what you are actually trying to communicate: Scientific Process, Subject Matter Fact, Emotional Connection, ...?