Traditionally it was assumed that using Captchas was a foolproof method for keeping out spam bots and other malware from having a field day on web based forms. This assumption may have been the Achilles heel in the adoption of this technology.
At the time, the idea of using a Captcha to secure web forms came into play -- the technology to beat it was not yet there. Probably the only problem was the usability of the Captcha; was the human user actually able to read the mangled contents of the Captcha? If not then it beat the purpose of the web form.
Over the years, as the potentially destructive technology and the algorithms driving them improved, they found a need to make the Captcha text more complex for machines to read, but fairly easy for humans to read. The problem with this is that there are always limits to how far you can mangle the image context without making it impossible for humans to read. Unfortunately Captchas do a great job of keeping out a large number of humans out as well as spambots.
With more and more services such as online banking being made available online, security had to take a paramount place in requirements and design planning.