Comparisons between playing live music and presenting science to an audience
Here are a few things that are similar and not similar.
Demos are like pilot experiments/pilot data
I'll have a rough outline of a song and I'll be unsure if it's a piece of a song, or the key to a song, or perhaps just something I made up and will throw away. That's not so different from science. I'll pilot something to see if it works and sometimes it does, sometimes it gives me something back I didn't expect and need to build around new expectations, sometimes it crashes and burns and I have to start again with something else.
Shows are like presentations (either posters or slideshows)
I come in with a setlist of rehearsed music. I set up on stage. I have a time slot and my audiences vary from show to show (sometimes big, sometimes small; responses are sometimes good, sometimes not so good).
In terms of posters, each section of my poster is like a song. And what I think is the most memorable piece of my set (the last section where I conclude and put findings in a broader perspective) may not always be what other people think is the most memorable piece (perhaps the task design or statistical approach). Depending on the conference, I'll have different audiences leading to different questions and perspectives on my research. Slightly different since unlike a show, I'm usually repeating myself as new people come to hear about my research -- like I'm playing a special set for maybe one person every few minutes.
In terms of slideshow presentations, I try to treat slides like pieces of songs while having as few slides as possible to represent a concept (where a concept I want to convey is most analogous to a song I'd play live). Most perceptions from poster sessions apply -- but this becomes more like a concert set since I'm usually on a stage in front of a bunch of people. To be honest, it's a very similar feeling.
Recording is similar to writing up a manuscript for publication
When I have all my demos talked out and perfected on stage and I have a feel for each song and can think about them in different ways, I try to record a set of new songs into some album (may it be an EP or an LP). I go back and forth with a handful of musical friends I respect to adjust certain levels or instruments or effects on each song for each instrument. It becomes really tedious and sometimes difficult -- especially when you enjoy some things but everyone is telling you "this doesn't work". Sometimes you take a risk and people really don't like it -- or people freaking love it. Makes the process more enjoyable when you can get past your own ego and produce things with others that is both enjoyable for yourself and for the consumption of listeners.
After a pilot experiment, I run the larger experiment to collect data in order to show that my initial hypothesis has quantitative support. When data supports my hypothesis, then sometimes a peer-reviewed paper is warranted. Sometimes they are short (like an EP) or longer (like an LP). I usually send it out to peers in my field first, to get a vibe for what needs editing and where.
Unlike the recording process, when you submit to a peer-reviewed journal, you don't have a lot of say as to who your reviewers will be (you can suggest reviewers, but you can't necessarily say I want so-and-so to review my stuff only). That's different since when you're recording music, you want specific people as producers in order to shape your sound (e.g. you want someone familiar with pop if you want a pop album, if you want to sound like Metallica then get Rick Rubin as your producer, etc). But ideally, your concept and your work should come through, regardless of who your reviewers are. Ideally. Just like how ideally, my music should sound the same regardless of who is producing songs with me.
There does exist "overproduction" in music -- that's not untrue of science either (same goes with "underproduction"). For me, I enjoy a community process with my music, but I am selective of who comes into that process. In science, I don't have enough experience with the publication side of things but for everything up to that, it's not so different (being selective as to who helps shape a manuscript with me).
Touring isn't dissimilar from attending conferences
I usually try to play shows in cities where I know people -- makes it easier to shower and have a place to sleep that isn't inside my car or costing me money. I have to eat out, of course. I have to be professional and on time. I have to set up and break down -- I have to do soundcheck and engage/network with people at venues. I am usually doing things at hours that I normally don't do things (e.g. driving at night, eating when I can as opposed to eating at certain times, etc). When I get back home, I'm exhausted.
Conferences are much like this too, except I usually only present one or two different days (versus every night, or nearly every night like on a tour). It's usually expensive (just like a tour), so keeping things as cheap as possible is necessary (e.g. conferences where I know people helps out a lot). Usually I'm at conferences for networking and engaging -- just like shows. I'm also eating with others who all have different schedules, so I eat and go out frequently so I don't miss folks I want to meet up with at conferences. And when I get home from a conference, I'm perhaps more exhausted than when I'm out on the road for two or so weeks.