Let's modify your question slightly: "Is it difficult to make MIPs for target molecules that are selective, robust, and cost effective?" The answer to that question is a very big YES! If it were easy, there would be a flood of products on the market. For example, I know a company whose charter is to develop MIPs to use as detectors of rather low concentrations of specific molecules that are produced in the body that indicate specific disease conditions while they are in an early stage and treatable. Also, some types of MIPs are being used to detect explosive compounds in soil and water, such as in areas of mine fields and abandoned industrial sites that made munitions for military use. A big problem with explosives is getting discrimination between the various types. You'll find many papers on this if you search for TNT, DNT, RDX, etc.
The "quick-and-dirty" method is to mix your target molecule into a reaction mixture and form a polymer, then wash out the target molecule and you have your MIPs. This will leave some of the target molecules locked in the polymer no matter how much washing you do. When designing your system, you need to consider how you will be detecting that the target molecule is in place in its cavity. Then you'll need to run several possible interfering molecules to see if they also give a response, and also if they fill the cavities and block your target molecule. Also, how long does it take from the time the MIP is exposed to the test material until a response is generated and stabilized.
Let's modify your question slightly: "Is it difficult to make MIPs for target molecules that are selective, robust, and cost effective?" The answer to that question is a very big YES! If it were easy, there would be a flood of products on the market. For example, I know a company whose charter is to develop MIPs to use as detectors of rather low concentrations of specific molecules that are produced in the body that indicate specific disease conditions while they are in an early stage and treatable. Also, some types of MIPs are being used to detect explosive compounds in soil and water, such as in areas of mine fields and abandoned industrial sites that made munitions for military use. A big problem with explosives is getting discrimination between the various types. You'll find many papers on this if you search for TNT, DNT, RDX, etc.
The "quick-and-dirty" method is to mix your target molecule into a reaction mixture and form a polymer, then wash out the target molecule and you have your MIPs. This will leave some of the target molecules locked in the polymer no matter how much washing you do. When designing your system, you need to consider how you will be detecting that the target molecule is in place in its cavity. Then you'll need to run several possible interfering molecules to see if they also give a response, and also if they fill the cavities and block your target molecule. Also, how long does it take from the time the MIP is exposed to the test material until a response is generated and stabilized.
It is difficult to develop protocol for good imprinting - with high specificity and low heterogenity of the binding sites. But if you have the protocol it is easy to reproduce it.