Still the same ideas as a year ago:
Public vote recommendations instead of dReps: Everyone can publish a whole slate of recommended votes, with or without short reasonings, with or without a general policy for those recommendations.
Using a standard format for those recommendations would allow users to import and combine them from multiple people whose opinion they value (but not necessarily unconditionally trust). Developing tools that just highlight the proposals where the recommenders differ would than be very easy.
If we finally get an API for submitting our votes instead of that horrible voting app that should have never gone into production, those tools could also allow to directly construct and submit a final ballot out of the recommendations, the decisions on diverging points, and the voters own preferences on some proposals.
Voters can then very freely decide how much effort they put into that. A fire and forget like with the dReps is totally possible, but if the possibility is there to at least have a look, to easily modify with own impressions on some projects known personally by the voter, to just import another recommender, my hope would be that a lot of voters at least do a bit of that.
It just needs an API to really submit the vote (otherwise that would still have to be done by hand). The rest is totally permissionless. We don’t need no freaking registration of very important dReps with some IOG bureaucracy that never answers. Everyone can just publish recommendations and voters regard it or not.
Change the voting system to highest-median: To remove the quite heavy strategic consideratrions of score or approval voting (what we have now, “Should I vote for that even if it could take away funding from my favourite?”, “Should I downvote it for that risk even if it is somehow okayish?”), switch to, e.g.:
We grade proposals with grades like “Bad”, “Passable”, “Good”, “Very Good”, “Excellent” and they get the median grade – the grade where 50% of voters gave a better and 50% a worse grade.
Since the median and not the average is chosen, there is no need to exaggerate.