Let’s take an example:
The proposal “Ouroboros-mini query specification” in “Open Standards & Interoperability” of Fund 8 got 22,475,160 ADA YES and 17,435,670 ADA NO. It only met Threshold #1, because of the NO votes. People voting NO, people who did not want to see this proposal funded helped it getting funding. If only 3,103,060 of those NO ADA had just not voted instead, the proposal would not have been funded.
If I participate in a vote, I don’t want to cause the exact opposite of what I am expressing in my vote!
There are voting system paradoxes that are hard to avoid. This is not one of them. It is just a threshold. And such a threshold should always be a threshold on YES alone (or on YES-NO), but never on YES+NO.
Or the other way round: Could you elaborate, why a NO vote should help a proposal get over this threshold? Why should the expression of opposition to a proposal help it?
(This does not affect Threshold #2, because in Threshold #2, NO votes already make it less likely that a proposal gets funded. The two thresholds are quite independent of each other and work on different measures. It’s totally okay to require a certain margin in percent by which YES has to be more than NO – arbitrary as such thresholds are always a little bit arbitrary, but okay.)
EDIT: Perhaps a direct comparison inside the same challenge helps even better illustrating the problem:
The “Codesign” proposal failed Threshold #1, despite it having significantly more YES and a much better score. If just 1,217,969 ADA more had voted on “Codesign” (even NO!), it would have gotten funded.
“Codesign” has been punished for getting less NO votes (much less) than “Ouroboros-mini”.