I'm not sure what you mean?
So, you could try the simple fix: a grunt fund calculated on the simple assumption that all hours worked are made equal. Say one person does 60 hours, the another 40, and you get a 60/40 split. BUT: the person doing the 40 hours was working on a patent application, and the person who did 60 was just bashing out template HTML. Not only that, there's a third person who's fronted some cash for a lawyer, and they don't get anything because they're working 16-hour shifts at their day job and haven't put any actual hours in. The shares look out of whack, and people get annoyed.
Next - and this is where my impression of the grunt fund idea comes in - you define an exchange rate between the various types of input up-front, so a 1 hour working on a patent application gains equal share as 5 hours writing HTML, or $100 paid out of pocket. Everyone is happy, everyone gets a share they feel is proportionate to the value of their input, and everyone's working towards the same "pay-off" day. The question is, what happens when it turns out, say, that they don't need a patent after all? The patent-writer's work has no value at pay-off day, except in a certain nebulous "we had to go through that learning to get here" way. Not only that, but the HTML-wrangler now has an incentive to get the company to pivot into that situation so that there will be no more patent work, and lots more HTML work. That's a weird incentive.
That's what I mean by the trade-off: in the first case, you have total uncertainty over who's actually going to do what. In the second case, you have uncertainty over how many hours people are going to work. In the third case, you have uncertainty over the value of those hours at pay-off day or, to put it another way, you have uncertainty over the accuracy of the exchange rates you define at the start.