Here is another neglected approach:

There is a research startup in ergonomics of logic programming that aims for augmented cognition while still relying on traditional interfaces with the computer.

More on this in my article: Become a Superintelligence Yourself

• The filter moves a lot of air, so you need quite long pipes.

Speed matter only in the need for bigger pipe diameter. Impedance depend on the ratio of length to diameter. 

• You need pipes on both the input and output

Indeed.

• Some of the noise will be vibration of the purifier body, so you might need to enclose that too

Depend on the quality of the fun bearing. Vibration of the air does transmit to the body and to the pipes, but very littlle. And if you can make an extra casing around, if you need.

This is how I was doing DIY purifier (first version of it) with a fun in the middle of pierced steel 60mm pipe. Wet charcoal on gauze wound around the pipe. Inside the pipe was a layer of reticulated polyurethane foam acting both as a filter and sound absorber. The thing afterwards put into 150mm pipe with a layer of polyurethane foam.

And that was quite quiet. The problem was in the quality of the charcoal giving a smell. So I reassembled it for better throughput and use in the kitchen without muffling.

And for the room I use this ugly and loud, but powerfull setup for short periods of cleaner air outside:

Why not? If it has a good fun, the most of the noise will be airborn, and so you only need to attach mufflers to inlet and outlet. But perhaps there can be difficulties if you didn't want to disasseble it.

I did for myself DIY purifier and used extremely noisy high-pitched 40mm fun for server units at 18000rpm. High frequencies even better get absorbed by the muffler. And due to the small size, these fans can produce high pressure at slow troughput. So can save on filter materials. Although didn't finish it since activated charcoal I bought was of low quality. Using it in the kitchen only. If only I could insert pictures in the comment. 

If you don't care much about compactness, you can easily make a DIY muffler: A long pipe with a sound absorber layer. Sound wave goes along the pipe. Its outer part gets absorbed and due to the diffraction inner part spreads. The impedance is exponential, or in decibells is proportional with respect to the length of the pipe.

It feels pretty inefficient to me. Filters are too big a resistance for air flow in this case and it will rather swirl around fan blades.

Why not attach filters right to the fan blades or even instead of them?
 

Also, academic research is funding languages like Haskell, Scala, Lean, etc.

I don't know about others, but Lean was funded not by academics. It was funded by a private fund ConvergentResearch. They target somewhere between academics and VC. I know this because Lean is the closest analogy to my project, although with a different mission. I tried to apply, but they didn't reply. Maybe they will find your proposal more appealing.

Anyway, the emergence of such funds as ConvergentResearch and OpenPhilanthropy is a promising trend.

Fellow russian researcher here. I doubt small donations can work. Not many people use crypto (I'm same noob and not sure if it's worthy to learn it until I have big audience.) I tried to register with OpenCollective so that they would collect money for me before I leave Russia but they rejected despite me having a project to showcase.

If you have something to show, you could try applying for research grants. This is what I'm trying to do. But that only makes sense if you have time and not so desperate about nearest future.

If those stories were about pursuit of truth, like Archimedes's eureka in a bathtub, they could motivate students and teach some lessons of rationality thinking.

And the history itself could be much more interesting subject if it were teaching some real wisdom rather than demanding stupid memorization of dates and places.

I do this all the time. But perhaps I should stay away from this conversation. Bad for the karma.