I think your whole logic is flawed.
The device won’t know when the value in the Blynk app has changed, because the device will be disconnected from Blynk whilst it’s in Deep Sleep. It will only know about the widget change when it wakes-up again, reconnects to Blynk and executes BLYNK_CONNECTED() and Blynk.syncVirtual()
This will mean you have a lag of up to 10 seconds before your device knows about the widget change. In reality, waking-up and connect8ing to WiFi then Blynk, then executing this code will probably take 10 seconds (although you have a timer set to 500ms to initiate Deep Sleep, and another one set to 1000ms to take some readings, so the readings will never happen)
That will give you about a 1:1 seep/wake ratio, which will mean your battery won’t last long. Extending the sleep time will improve the battery life, but make the servo even less responsive to widget inputs.
I wouldn’t use the Edgent sketch for Deep Sleep. You could add the Blynk.Air functionality to a non-Edgent sketch using this functionality…
But you need to ensure that the device doesn’t go to sleep while an update is taking place, which can be tricky.
Also, you shouldn’t use timers in a Deep Sleep sketch. You don’t want to wait around, you just want to get on with the operations in the desired sequence then go to sleep again.
This topic is rather long, but you’ll read a lot about suing Blynk with Deep Sleep if you study it…
But, we come back to the same problem, which is that you cant use Deep Sleep if you want interactive control over a device via the app or web dashboard.
One other thing, when you post compiler error messages or serial output in future, do it with triple backticks at the beginning and end (the same as when you post code), not Blockquotes.
Pete.