Recently, a shareware author wrote that one can no longer sell traditional shareware. What he meant was that successfully selling shareware products now requiring a level of marketing and development effort that is becoming indistinguishable from that required by full software companies. Perhaps, there are too many applications available now competing for buyer's attention and dollars.
I came across an article that suggests that shareware can be profitable. In this post on why people register products, The author claimed that crippling software does in fact lead to substantially higher registration rates and revenues. He performed an experiment by writing an application, Smart Doc, in a few days and releasing it to the web. The software used a random number generator to decide with a 50% chance whether the application installed would be "crippled." The application generated $34,075 after one year of release. Had all copies of the software been crippled, the revenues would likely have gone up by $17,125 to $51,200.
It was enlightening to see that a few days of effort could generate over $50K, but this was back in 1995.