No, it’s ‘black magic’ (if we’re allowed to use that term anymore) in the BlynkSimpleShieldEsp8266.h library file that is included in the sketch examples that use an Arduino + ESP8266 shield.
When ESP8266’s ship from the factory they are programmed with AT firmware. When you uploaded a sketch to the ESP you over-wrote this, and if you want the ESP to act as a WiFi modem for your Mega then you need to re-install the factory AT firmware. This can’t be done using the Arduino IDE, you have to use something like ESPTool.
Then, the ESP8266 (which incidentally is several times more powerful than the Mega) will act as a dumb modem for the Mega and give it a limited degree of WiFi functionality.
The board you’ve chosen is probably the worst possible way of getting started with IoT connectivity (and ridiculously expensive for what it is as well). Getting the switch settings right, and using the correct serial port configured in your sketch is confusing for beginners and at the end of that process you still have a low powered MCU with no IoT connectivity which is using a higher powered MCU as a WiFi modem. Something we tend to compare to the idea of using a sports car to tow a cart.
Admittedly, there are some situations where an Uno or Mega with an ESP8266 are better choices than an ESP8266 on its own, but often in those situations an ESP32 is a better overall choice of board.
I’d usually encourage newcomers to start with an ESP8266 based board such as the NodeMCU or Wemos D1 Mini, as they are much simpler to get up and running with, and don’t have the drawbacks associated with the Uno/Mega plus ESP8266 combo. If they then need to use an Uno or Mega based solution then what they’ve learned with the NodeMCU will be an excellent grounding in IoT systems such as Blynk.
Pete.