I got a cat 3 years ago, it was not viable, it still isn`t cost effective, yet it was worth it.
Regarding that CPU thing, if your awesome electrician never tried to build a CPU because someone told him that it isn`t worth the time, you`d be off with the $150 one, wouldn`t you? His first project probably blew smoke, or was crap, and had to throw it away, did that work go in vain? In my honest opinion experimenting is always beneficial, regardless of the outcome (if it doesn`t hurt the project`s - if any - professionalism, that is).
Also, there is prototyping, a damn important part of a lengthy development (like a full-fledged webserver perhaps?).
Regarding that CPU thing, if your awesome electrician never tried to build a CPU because someone told him that it isn`t worth the time, you`d be off with the $150 one, wouldn`t you? His first project probably blew smoke, or was crap, and had to throw it away, did that work go in vain? In my honest opinion experimenting is always beneficial, regardless of the outcome (if it doesn`t hurt the project`s - if any - professionalism, that is).
Also, there is prototyping, a damn important part of a lengthy development (like a full-fledged webserver perhaps?).