It seems that enscript is not able to correctly map the spaces in the text file that it receives as input even if the afm files generated by ttf2ufm contain the entry for the space code (32) (see the last comment in the article). Any help on this problem will be highly appreciated!