All rewards are funded by a "tax" (i.e. inflation) on stake. I'm not sure where else you think they come from.

Unless you are proposing to further increase the already-high 8.5% inflation (in which case the answer is very likely to be a hard no from nearly all stakeholders), then the only way this can be done is by reallocating existing inflation-funded rewards. For various mathematical reasons starting with the Willie Sutton Rule ("because that's where the money is") most or all of this reallocation has to come from author rewards.

I can respect if people think that author rewards specifically are inherently more important and valuable to Steem then a flexible budget system for funding proposals, though I disagree. Author rewards IMO are one way of adding value to Steem by attracting a user base and supporting engagement, but not even remotely the only way. I'd rather have a budgeting proposal system which can decide among competing priorities in how to spend the budget to best help Steem. That could, conceivably include giving the funds right back to authors, if stakeholders (i.e. the ones paying for ALL of the rewards) really believe that is the very best possible use, or it may include other things (marketing, development, promotions, referral programs, etc.), some or all of which could and likely would directly benefit users, including authors, albeit in different ways than direct monetary payment. I would contend we have been sorely lacking in the latter even while more than generously funding the former.

There is more that needs to be done for Steem to be a success than just paying out most of the inflation budget to authors, otherwise it would already be doing what needs to be done and would be on a clear path to success. But I don't believe that to be the case, and I doubt too many other people believe it either.

Finally I would reiterate what was said in the original post: that a significant portion of "author" rewards are currently being used to reward various ongoing projects, community groups, development work, initiatives, apps, etc. and rightfully so. But to the extent those things can shift over to getting longer-term funding via a proposal system, that removes those costs from the regular reward pool, leaving more for regular "authors". I seriously doubt that the net reduction to these "regular authors" would be anything close to 20%.