Look re digestion was common technique used when sequencing technique was not fast. U cud get fragments of DNA and people used to do mapping. But today with advent of deep sequencing it is easy to have whole sequence rather than having maps. People now a days dont use this technique. It is now an out dated method.
If you are trying to identify the microorganism then using the well established primer sets for 16S rRNA which should give you a product you can sequence old generation locally. If you have a population of microorganisms then NGS metagenomics is the way forward - again there are plenty of such papers from which you can read the materials and methods to get an idea of how people prepare samples for sequence analysis.
See https://www.ebi.ac.uk/metagenomics/ or http://metagenomics.anl.gov/
well to save money I would say you could do it. But you have to use enough restriction enzyme to have a good resolution. Meaning that if you just use one enzyme you could have with the same restriction pattern sequence which are affiliated to different species. This means that you should use a set of enzyme, I used three. So for each clone you will have three restriction pattern, of course this complicate a bit you life if you consider all the possible combination.