I would like to give some advice to people making kaizo hacks. Watch this video and, in particular, note how often the player dies on the very first jump.
http://www.youtube.com/watch?v=56LfMCXDGw4
The player is an extremely skilled player, good enough to beat all of kaizo mario without tools, and he had about a 40% success rate on the very first jump on this level. Now, if you let an average SMW player try this jump over and over again he probably won't have a much lower success rate than the extremely skilled player from the video. This is because the jump requires you to jump at a very particular time and no human is capable of detecting exactly when the right time to jump is. So essentially, skill is not a significant factor for success at this jump, it is mainly just a matter of luck.
So now imagine an entire level of jumps like this that require levels of precision that no human is capable of achieving and what you have is a level in which skill doesn't affect success, that is, a really bad level. And so I think there are two types of hard: "good hard" and "bad hard". For "good hard", skill will be the main factor determining success. For "bad hard", luck will be the main factor determining success. Most of the original Kaizo Mario I would characterize as "good hard" with a few exceptions, this jump being one of them.
A lot of people who make kaizo hacks seem to think that the only way to make a level hard is to put munchers everywhere and rely mainly on precision jumps for difficulty. I think that this is the easiest way to make a hack hard but unfortunately it will usually end up being "bad hard". There are other ways to make a level hard that don't rely entirely on these things. I think some munchers are usually necessary to achieve high levels of difficulty but it is by no means necessary build entire levels out of munchers.
So I guess what I'm saying is ease off the munchers and the precision jumps and try to make kaizo hacks that are "good hard" and not "bad hard".