Aside from manufacturing considerations, there are also software related issues. More PPI mean that much of the stuff that you see on the screen will get smaller. Software developers have been addressing that issue through resolution independence, but resolution independence isn't universal because of the data involved. (One such example are rasterized images, such as photographs.)

Another hitch is a performance hit. More pixels mean that computers have to deal with much more data. Going from 100 PPI to 300 PPI involves 9 times as much data being pushed around. In many cases you also have to deal with rendering as well, though I don't know if that goes up with the square of PPI.

I do agree, higher PPI monitors would make life much nicer. That's particularly true if it brought about the end of anti-aliasing where contrast is important (e.g. text and line art). Yet I would caution anyone against thinking that the solution is simple and inexpensive. Monitors like the IBM T220 were very expensive and intended for high end workstations, workstations that ran specialized software that spent more time doing actual work than the whiz-bang stuff we expect on personal computers.