Other than not being practical for a number of reasons - I'm not going to delve into fanciful 'designs' that again, aren't in the least EVER going to be practical ... synthetic aperature RADARS work well when you can saturate a target with pulses, taking a trade-off in other parameters; parameters you don't have 'to trade off' from a platform in geosynchonous orbit ...
I sense a total lack of appreciation for this 'field'; received signal strength (and transmit ERP) for *any* system is directly proportional to antenna gain - a factor which is directly controlled by 'capture area' which relates to an antenna's size ... the ONLY way to make up for a signal which has 'sunk' below the noise floor (intrinsic noise in the first active element e.g. as a front end preamplifier) is to increase antenna gain via an increase in it's size there by affecting the S/N (Signal to Noise) ratio ... even so-called Pulse Compression 'chirped' systems which can function with a N'S ratio require a workable N/S ratio prior to 'pulse compression' (in the receiver) and the resultant S/N improvement!
There is no cheating nature and the 'fact' of thermally induced/created 'noise' in all active devices (and even passive devices!) as used in any RF/RADAR/Radio system!