Fact check: "The horrors that come with methamphetamines"

Checked from background knowledge, not a fresh literature search. Figures are approximate.

Verdict

The core is accurate. The overdramatisation is mostly in the generalisations. It is credible first-hand testimony from an extreme context (long binge, street life, trauma, heroin history), presented as what meth does to everyone.

Holds up

• Meth psychosis is real. Studies of regular users, including Australian ones, find roughly a quarter report psychotic symptoms in a given year, and they are several times more likely while using than while abstinent.
• Sleep deprivation plus meth is the main driver. Trauma history and a life of constant vigilance raise the risk. Paranoid delusions of being hunted or shot at are typical.
• "Productive, then stuck" fits the known pattern of repetitive, pointless tasks (called punding) after long use.
• Running down the street naked is plausible, not a flourish. Meth raises body temperature, and overheated people often strip. Add skin-crawling sensations and a delusion of being hunted, and bolting from a building half-dressed or naked is a recognised presentation. Beating on doors fits the same picture: he believed the danger was real and was looking for shelter.

Overstated or shaky

• "You don't withdraw" after switching from heroin to meth. Meth does nothing for opioid withdrawal. Acute withdrawal peaks within days and mostly passes in one to two weeks, so at the two-week mark she had likely just come through it.
• "Four days awake and you will see demons, hear voices, go into full psychosis." By day three or four, sleep-deprived people commonly get visual distortions and paranoia. Voices and full psychosis are not guaranteed, and Randy Gardner's 11-day stay-awake record produced hallucinations and paranoia without lasting psychosis. With meth on board it is far more likely. True as his experience, not as a universal rule.
• "Traumatise them the rest of their lives." Most episodes clear within days to a couple of weeks of stopping and sleeping. A minority persist or recur, and in some large registry studies roughly a fifth later get a schizophrenia-type diagnosis. The memory of the episode can stay traumatic.
• "People talking to themselves are living in a nightmare." Talking to oneself has many causes (phone earpiece, mental illness, other drugs), and not everyone experiencing psychosis is terrified.
• "Hear a cricket fart a mile away." Figure of speech. Hypervigilance sharpens startle and noise sensitivity, but not literally.

Why some people stay up for days, repeatedly, without psychosis

Staying awake longer than the speaker did does not by itself predict psychosis. The risk depends on more than hours awake:

• Individual vulnerability. Genetics and family history of psychosis matter a great deal. Most people who use stimulants heavily never have a psychotic episode.
• Trauma and living conditions. The speaker had street life and a trauma history, which prime the brain for threat. A stable setting with food, shelter and safety lowers the odds.
• Dose and pattern. Risk climbs with dose and with continuous redosing. A planned weekend with a ceiling on intake differs from an open-ended run.
• Which drug. A long night awake on a different drug is not the same exposure as days of meth.
• Recovery. Sleeping properly afterwards resets the system. Running straight into the next binge without recovery is what compounds the risk.

One caution: no psychosis so far is not immunity. Repeated exposure can sensitise the brain over time, so the threshold for an episode can drop, and an earlier episode makes the next one easier to trigger.

Bottom line

Swap "will" for "can" and "often" in the original claims and most of the overdramatisation disappears. His message to be kind to people in that state is sound.