It's a simple, elegant approach covered by quite a few papers. If it has a downside, it's that you do need to _know_ what the "maximum power voltage" parameter is for your panel. Maybe this can be acquired in production line calibration. Solar panels are rather variable from batch to batch, so you can't just fit and forget this sort of electronics to an arbitrary panel. The extra complexity of a routine that goes looking for the maximum power point does reduce this concern. but for 'emergency' operation, the last few percentage points of conversion efficiency aren't your main concern.
Great article with a host of practical circuit suggestions. I can see using this approach on a handheld battery-based instrument used in the field. A solar charging option is a perfect feature when you can't get back to 'base camp'.
Replay available now: A handful of emerging network technologies are competing to be the preferred wide-area connection for the Internet of Things. All claim lower costs and power use than cellular but none have wide deployment yet. Listen in as proponents of leading contenders make their case to be the metro or national IoT network of the future. Rick Merritt, EE Times Silicon Valley Bureau Chief, moderators this discussion. Join in and ask his guests questions.