I think you have a rather odd view of computer history.
Back then computers were much more limited so it was possible to do something revolutionary with a relatively small numbers of people and relatively small amount of money. Doing the same today - if it could even be done - would cost Billions (not millions) and require an army of highly qualified people.
As for competition being less, you are kidding? right?
You'll be going up against a massive installed base of pretty much one one highly entrenched system. It's much, much easier to go up against multiple systems in an immature market.
You're basically talking about developing a chipset significantly better than AMD/ATI or Nvidia could develop and producing an OS that would then easily sweep aside Windows - as well as the rest of the PC industry.
You really think that today would be easier than doing it in 1985?
To put it into perspective, the leading PC architecture in 1984 was the Commodore 64.
So ok, we all roll over and die then.
BTW it wouldn't be that hard doing better than NVidia and Ati lately considering how their overpriced card consumed more power than my space heater and they are so buggy as to necessitate weekly drivers updates.
It would be easier today. The tools needed to design them are more powerful today than they were in 1985, and engineer are better trained today in that field. The manufacture plants are also better and more efficient.
God, with attitude and views like that we would have never made it out in space.
As for the entrenched base statement, ask Apple about it. Nobody would have given them a chance in the late 90's but look what innovation did to them. You have to take risk if you want to progress.