My SR question about configurable VPNs did not have much response over 8 months, so I added a hefty bounty to attract attention, which it did.
None of them was an obvious best, but most of them seemed to provide pretty good feedback.
Unfortunately, there isn't enough time in the bounty window to actually evaluate the answers by trying them, especially as some of the answers arrive late in the window. The bounty period expired and I had to award it, so I used my best intuition and chose one.
But that award comes just from guessing. I don't know which answer is the right or best one. Nor do I feel that my awarding the bounty should be interpreted as a "this OP recommends this answer" but a big bounty awarded is surely interpreted that way by other readers.
That seems pretty unsatisfying. The respondents all did well, and I would liked to have distributed the bounty across the answers but there seems to be no way to do that.
I suspect that SR answers must often be in this category: interesting "recommendations", not enough time to evaluate which is best if a bounty is awarded.
Feature request: enable OPs to split a bounty across multiple answers.
For reference: the same request on the main meta, which was declined. The discussion there revolves around the notion that the bounty should be awarded to the obviously best one answer. My point here is that there is not yet such an obviously best answer, making the premise leading to the conclusion false IMHO
Feature request: enable OPs to move an awarded bounty, when a particular answer has had time to be evaluated as "best".