I don't think you will find this information in a book. Re-crystallisation is a skill which comes with practice.
Unfortunately if you want good quality, decent sized crystals from samples which are notoriously hard to crystallise then patients is your only option. The more rapidly you form crystals the smaller they will be.
Depending on your compound hi-vacuum sublimation onto a cold finger is sometimes an extremely useful and often overlooked technique for growing X-Ray quality crystals, from small organic molecules, in a relatively short period of time. Using this technique you don't form solvates either which can be an advantage in some circumstances.
For your information though accurate and useful elemental analysis data is far more reliant on purity of sample being crystallised than single crystal X-Ray.
Unless you have vast amounts of impurity which is going to co-crystallise with your substance then provided you have a relatively pure sample your crystal will be of sufficient quality for X-Ray. Any impurities on the outside of the crystal lattice which are disordered do not interfere with the analysis. I have successfully used samples which have been crystallised from very impure crude mixtures.