Oh dear,
Great question! I wanna say first that several members are indeed helping CT, via patreon, and I respect that so much. I really don't take that for granted and am grateful for the show of faith, and can only hope to reciprocate that back. But yes, there's still so much room for creative involvement in CT -- to help it expand more rapidly.

But I’m not exactly sure what we could do to be most helpful. So I’ve been wondering, what do you all think would be the most useful/impactful thing

Hmm, I don't know if there's a way to answer what is most helpful-- since there are many different angles to tackle it from, depending on what talents a member has. I think the most helpful thing would be whatever a given person can contribute to this exploratory journey.
But maybe I can list aspects of the vision that I see for CT's near-term future, to give some jumping off points?

• Host new Vultology Quiz: Technically, if anyone wants to organize another vultology quiz like this one, they can do so. I am a participant in it like anyone else, and the point of the quiz is to see if we can achieve ever-better levels of inter-rater reliability. Last time we got ~90% consensus among the top four, and ~80% for the rest. That is excellent but to reach higher scientific standards, I think we can reach ~95%+ consensus. Anyone up for hosting it? 🙂
• Research Demographic Bubbles: This is a nice way to add samples to the database while exploring an area of interest. The idea is to pick a demographic bubble (i.e. top athletes, presidents of every country, someone wanted to do one on tarot/crystal witches) and type everyone relevant. Just start a thread, and paste all the samples you find. Ideally using the codifier to write up a report for each. This won't be a "controlled" selection, but we're not quite yet at that point, and first-passes like this often reveal a lot of new things. You'll be surprised what vultological parallels there are within humanity. o.o
• Reach 1,000 Public Database Samples: Speaking of the database, we've looked at several thousand samples by now, but not all have been added to the database. The public database is currently hovering around ~580, but now that the codifier's done this can easily pickup in speed. This is key, because it's the backbone of the model and what allows meaningful statistics to arise. Does anyone want me to make a tutorial video on how I add samples to the database? If so, I'll make one.
• Run Statistics: Members like @supahprotist and @ladynerdsky have been great at this, and you too fayest! But there's so much untapped potential. The database is an untapped source of insights. A good case study for something that could come out of this is another career analysis breakdown. Back when the old pilot study was done, we didn't have development levels. It would be interesting to see how this affects statistics too.
• Social Presence: This is something I'll be doing soon as well, by making youtube videos. What would be especially helpful is a series of videos that can help beginners ease into the CT model by answering a lot of the basics. Anyone else interested in being a ytube host? Maybe we can also have podcasts and discussions about important topics.
• Research Hereditary Ties: There's also a pending need to look into the hereditary component of CT more closely. This could be done either by typing one's whole family (get a vid of each family member 🙂 ) or finding several celeb families to look at. So far the pattern shows that there is a functions quadrant hereditary component, but energetic quadrant is random. But if this could be further verified (or denied) it would be an important theoretical point to consider, regarding the reality of CT.

These are some of the things I can think of just on the spot, that are within reach + hopefully not too dry/boring. I'll save the grunt work for myself, as I feel bad about anyone else doing my grind for me. ;p However, I have loads of other recreational thoughts that could be useful thread-ideas. Such as:

Thread Ideas

• Type Collages: On books/movies/works specifically by a given type range or demographic. Such as those done in this old thread. Other people do collages on psychologically typed individuals, but it's another thing altogether to see the similarities in works by wholly visually typed people.
• Political Differences by Type: Naturally there's no 1:1 correlation between any political viewpoint and a type, but it's interesting to compare how political views distribute themselves statistically across the types, because it also isn't wholly even. There are differences between Revisers and Conductors, or Je-heavies vs Ji-heavies more generally. There are differences between Fe/Ti users and Te/Fi users. We can learn a lot from such discussions.

 
If any of this sounds exciting and interesting to anyone, and you wanna tackle it, I think a lot of unexpected insights could come out of it.