> ... how would you most effectively distribute them? 

Make a list of 100 interesting authors and check their blogs <em>manually</em> on a regular basis. That way you can make 40 of your 50 "impactful votes".
For the 10 remaining "impactful votes" try to find good posts from authors you have never read before to give new users a chance and some motivation to keep writing.

> And the modest bump to each one would not provide much exposure (e.g., on trending), which means it's unlikely for others to find it and place more votes on top (and increase your curation return).

I don't understand why investors always talk about their ROI (here in this special case "curation return")? Their main aim should be to increase the <em>value</em> of STEEM. Voting for 'quality content' (and flagging bid bot supported posts) would be a contribution to a higher value ... (with or without much curation reward).

The <a href="https://en.wikipedia.org/wiki/Network_effect"><em>value</em> of a (social) network</a> is measured among others by the number of its users.
So lets make sure that as many as possible users are having a pleasant user experience (for example also because they might get some impactful manual upvotes from time to time) and thus <em>stay here</em>. Lets do that as investors to save our investment.
If you have one million STEEM it's <em>not</em> most important to get even more STEEM, it's important to increase the value of <em>these</em> STEEM you already own.
For example I have much more STEEM than a year ago, but my account value has decreased significantly.

Did I need 'ROI' when I bought BTC some years ago? Or am I happy about the high <em>value</em> of my BTC nowadays? :)

Nothing against ROI, but I think we shouldn't be that focused on it ...

I wouldn't neccesarily decide that I had to give an equal vote to 50 post for one thing. I would give votes at different percentages based on what I thought they deserved. I do see this dilemna and that's why I'm not railing against delegation or curation guilds.  My points were mainly about bidbots. How do bidbots help to scale?

I definitely see how delegation and curation guilds can be justified for the reasons you mentioned. And I see the use in projects like @curie and @tribesteemup, but those don't make the entire trending page sold to the highest bidder like bid bots do.

>To find 50 good but undervalued posts on steemit.com every day might take quite a while

I don't know what you'd consider good but stuff like this deserves to go undiscovered not earning nearly $150.00  using bidbots to go trending.

https://steemit.com/art/@jellenmark/thanos-reimagined-with-infinity-2019-7-1-16-59-28

You guys actually think that people are going to downvote crap like this after the hardfolk to keep folks like this in line but the reality is when brought out into the spotlight in a post not even SFR would touch it because of who was behind paying the bidbots on it.  This post was at the most worth about as much as the paper it was written on.  This is one of the major reason Steem can't get out of it slump and attract new users.

Oh really, I didn't even realize it was originally 50/50. I hit 3 years here the other day and as far as I'm aware it's always been 75/25 during that time. I could be wrong about that as I'm just going off of memory.

I do think HF21 is a step in the right direction towards rebalancing those incentives. I'll be paying close attention to see the effects. I like the idea of incentivizing that altruistic behavior and I might even prefer that to instilling it as a culture, but I'm open to both routes to moving away from vote selling completely.

No it's not.