I have read with interest the exchange between the above two respondents. I tend to agree with both of you. And for whatever it is worth, here is my *"two-bit"*:
Firstly, I do have some actual experience of having handled all kinds of imagery - sub-metric [less than 1m resolution], metric [generically means 1-6m resolution], and low-resolution [10-meters or greater].
Real time when referred to in terms of satellite imagery can be confusing even for the practitioner. Unfortunately not much has been done to dispel the doubts. On the contrary, one gets the impression that the various schools merely lend credence to all these conjectures through their explicit silence on the issue.
If you consider the image taken by a sensor onboard a satellite, it resembles absolutely nothing anyone has ever seen before. This is the raw data. It could as well resemble a piece of real estate in the Saharan desert or even the Mongolian plains. The geo-coding is the first step in the correction process. Thereafter is the next stage of making the image viewable by one or more of the software available for commercial use. This processing of the raw data requires high computational systems [much of which are common place now], and conversion takes time - which can be speeded up to a certain extent but can never be made "real time". So if you are talking about 'actionable' inputs to be gleaned from a satellite image, it can only be undertaken after these two first steps have been completed. With more and more autonomous satellite payloads, much of these functions are being carried out onboard the satellite itself, but how much of such imagery would be made available to the public, is still questionable.
Next is the issue of accuracy of data in terms of map positions etc. If you imagine that when you run your cursor over a geo-refrenced image that has undergaone a first level correction, and you will be able to get accurate positions [approx positions - yes, but not accurate], then you are sadly mistaken. This requires one further iteration. Also, for this you need to have the proper projection system superimposed on your image depending on your observed latitude. A further addition to the delay in viewing "real-time" imagery.
Last but not the least: It is a fact of physics, and not hype by any community regarding panchromatic images being available in metric and sub-metric resolution. Yes, one can have a 'merged' product that has the resolution of a pan image, and yet the color details of a multi-spectral image - but that is done through human intervention & software - its not automatic. Till some more research is done I reckon we will continue to get only pan images of sub-metric quality.
Kindly do not compare UAV and satellite imageries. They are chalk and cheese. And yes, there is a time-lag between the image being captured and being made available to the final recipient. And a UAV gives inputs from different sensors. So to make holistic sense out of this input, one needs to undertake data fusion - so one needs to take care of data of different formats, different kinds and then fuse them to make a composite whole. You and me don't get to see this process taking place, but it nevertheless happens.