I personally am so eager to see this realized because one of those areas shown above is my dream place -:), but there would be many challenges to be overcome. Followings are a short list of those challenges. (NOTE : To foresee on those challenges, I think it would be helpful to investigate on the challenges that the satellite communication systems being depolyed and tested now (as of Feb 2021) and SpaceX Starlink can be a good example since there are so much informations are available. Even though it is not easy to find the details on challenges on the fancy looking system, it would be more helpful as an engineer to look for the challenges and see how those challenges gets resolved).

Latency : One obvious challenge that everybody can easily guess would be the long latency caused by pure physics. There is long distance between ground station/user terminal and the satellite. The speed of light is finite. So it would take a considerable amount of time (delay) for the radio wave to reach the user device. Of course, this latency would very depending on the altitude of the satellite. If the satellite is deployed on GEO(Geostationary Orbit), the delay would be huge. If they are deployed in LEO(Low Earth Orbit), the delay would be low (in some case it can be even shorter than what you experience with WiFi delivered by ground connection), but it would require huge cost to cover the wide area with LEO like StarLink.

Demand on Ground Stations/Integration with terrestrial core : Even though we can put the radio part (e.g, gNB, Relay etc) up in the space, the core network is (should be) on the ground. It mean that the satellite or airborne platform should be connected to ground station at some point. The question is how far it can get connected to the ground station. Obviously there would be a certain limit in distances in which a satellite can reach a ground station. As a result, we would need a lot of ground station as well as satellite if we want to cover wide area. In some cases, it will be very challenging to set a ground station. For example, if we need a satellite covering far in the oscien, how can we setup a ground station that can cover those satellites ? One possible solution for this situation is to make some satellites to get access to the core network indirectly via another sattelite instead of ground station. In order to do this, we would need to make relay networks or mesh networks among the satellites. Of course, theoretically this is possible and most of satellite communication systems put this into account from the design phase. But it would not be easy for implementation in reality (NOTE : This is one of the challenges that Starlink faces at early phase and as of Feb 2021 the communication between satellites is not supported)

Antenna Technology : Intuitively and by experience, we would easily guess that Antenna technology would ge a critical component for this kind of communication system. First how small we can make it ? We may tolerate the necessity of using a large sized dish antenna (i mean large size in comparison to mobile phone antenna), but everybody would like to have a smaller dishes in various reasons. Next, how to change the direction of antenna radiation pattern pointing in line of sight to a satellite ? Two options are considered for this. The first option is to put some mechanical component so that it can change the direction of the dish mechanically (like the dishes used in Starlink system). But it would be difficult to change the direction of the antenna fast enough to handle mobility situation. The second option would be to use phase array antenna by which we can change the direction of radiation pattern electronically. Technically this would be much better than the mechanical method, but cost would be a big issue.

User Terminal Alignment: For end-users, especially non-technical ones, aligning terminals to ensure optimal connectivity can be a significant challenge.

Large Doppler Shift :  Usually sattelites and airborne platform in this communication system moves very fast whereas most of user terminal is stationary or moves slowly. It implies there would be huge differences in terms of relative speed between satellites and user terminals. In turn, it mean there would be large doppler shift experienced by the reciever.

Large Delay Spread : In this kind of communication system, the physical distance between the transmitter and the reciever tend to be very far (e.g, ranging from several hundreds of kilometers(LEO) to 36,000 kilometers (GEO)). This would lead to large delay spread.

Competition to other technolgies  : Simply put, in terms of business point of view, would it be lucrative enough to invest huge money to compete against existing satellite communication system like Starlink or Kuiper ? (NOTE : There is a company called ASTspaceMobile that claim to provide SpaceMobile service with just a few hundreds LEO satellites for regular mobile phone. But almost no technical details are known as of now (Mar 2021)).

Handover complexity - Handovers between satellite beams/cells and between satellites and terrestrial networks could be complex and require tight coordination. Fast moving LEO satellites complicate this further. Handling handover may be easier (so advantage) in NTN comparing to conventional satellite protocol, but it still be more difficult comparing to terrestrial cellular technology.

Network Management and Traffic Optimization: Managing the traffic over a satellite network is more complex due to variable conditions and the mobility of the satellites themselves.

Power limitations - Satellites and aerial platforms have restrictions on power, antenna sizes, etc which can limit data rates and capacity. Efficient waveforms and modulation are needed.

Scalability: As the number of connected devices continues to grow, NTN systems must be scalable without incurring prohibitive costs or overly complex network management.

On-board Processing Capabilities: Satellites need to have sufficient processing power to handle complex operations, which can be a constraint given the limitations on size, weight, and power.

Launch costs - Getting satellites into orbit is still expensive. Larger constellations could require dozens or even tens of thousand of launches.

Orbit management - Avoiding collisions and managing constellation orbits takes planning, especially with large numbers of satellites. Debris is also concern when the collision happens or when the sattelites got out of life.

User terminal cost - For widespread consumer adoption, low-cost yet high-performance user terminals are needed. Striking the right balance on capabilities could be difficult.

Regulatory hurdles - Getting regulatory approval to launch large constellations and operate seamlessly across many countries can be challenging and time consuming.  

Business case - The large upfront capital costs make the business case difficult, especially when competing with expanding terrestrial 5G networks. Return on investment is a question.