Why almost no allin range on this river?

As per your graphs, trips have from 90% to 97% equity, and MDF vs a 4x shove is 20%.
If you add to the equation the fact that good trips hard block most of villains top of range, you end up with most trips not being able to shove for higher EV than a normal overbet.

Best trips probably could shove for an indifferent-ish EV to 150%, but it seems like you don't have enough of those to justify the line

I don't understand what the equity of the trips have to do with the MDF.

Hero range also have bluffs, and villain must defend the MDF to make hero bluffs indifferent.

You bet with value to make money when you get called. There's a size that maximises how much money you make and it has to do with your equity when called

I was confused because I know we have a shoving range in some spots with nut flushes, and they have 100% equity against villain range.

But reading you again, you mention we could shove the best trips but we don't have enough of those (we actually have a tiny shoving range with some of those and some 22, so that makes sense).

So you consider 90% to 97% equity actually not that good to shoving x4.

I started playing with a simple excel formula I made for equity, pot size, bet size and mdf, and I can see how we want smaller sizes with fewer equity as this manipulates the mdf.

I get that hands with 90% equity are happy betting ~x1.2 pot.

While hands with 95% are happy with betting ~x2.2 pot.

I guess this has to do with the impact of "traps" on bet sizing Michael Acevedo and Tipton's talk about in their books.

Hands with 100% equity are interested in betting the max in a vacum (there's the range protection factor in practice).

IP, it was interesting to find we need around 72% equity hands to bet 3/4 against an opponent not bluff raising and make it worth the same as checking.

While OOP, a blocking bet of 1/3 with ~62% equity hands is worth it around the same as our equity against an opponent not bluff raising. Of course now we also have to factor that checking doesn't close the action and we can get equity stealed by an opponent betting a balanced polarized range which incentivate more thin small value betting compared to IP.

I see hands with even around 50% equity blocking betting on the river oop for 1/4. This seems to vary some from spot to spot but numbers seems to be around there on average for borderline blocking betting hands.

Thanks for pointing this out!

You're a fast learner, I like it!