Thanks for everybody’s feedback!
The question remains, why should the multi-bar rest be split after the first bar? The musical information in the first bar applies to the entire silent passage. The current behaviour results in a confusing visual impression, forcing one to avoid multi-bar rests when using chord symbols over silent passages.
It seems to me that chord symbols in multi-bar rests should be handled in the same way as, for example, text or dynamics, already are. Text objects and dynamics only split a multi-bar rest at the beginning of the bar at which they occur, not after that bar. In particular, they don’t split a multi-bar rest at all if they occur in its first bar.
Pianoleo, in some situations what you describe is a great approach. You can then also hide the single-bar rests, to make it even easier to annotate the score. That said, I hardly use this approach anymore. For me it’s more important to keep the scores short and page turns to a minimum. At the same time, I prefer to have chords for the entire song, not just individual passages. Also, there is never time for clean annotations (especially when you’re on an iPad or tablet). When there is time, I’d rather go back to Dorico and update the score.