Explain the theoretical boundaries of lossless text compression based on Shannon's source coding theorem. Discuss how the entropy of a natural language corpus relates to the minimum possible average code length. Provide a mathematical derivation of how redundancy in text allows for compression, and discuss the theoretical implications of a text corpus having maximum entropy.