CAUTION: The following post (or a page linked to) includes or discusses loading data not covered by currently published sources of tested data for this cartridge (QuickLOAD or Gordon's Reloading Tool data is not professionally tested). USE AT YOUR OWN RISK. Neither the writer, The Firing Line, nor the staff of TFL assumes any liability for any damage or injury resulting from the use of this information.
In general, a crimp by itself isn't terrifically strong. In a rifle, we know the ultimate equivalent of a super-crimp hold is to seat the bullet into the lands. This raises start pressure about threefold and peak pressures about 20%. The next harder level rifle "crimp" I am aware of was the one year the National Match ammo had tinned bullets that cold-soldered to the case necks in the 1920s. Not only were there high pressures, but bullets were retrieved from the berm under the targets that had case necks still attached to them that had been torn off and squeezed down and through the gun barrel with the bullet. Now tearing the neck off is some kind of hard pull!
The bottom line from the above: for medium power rifle cartridges, there is simply no way a crimp can take pressure beyond normal tolerances. Indeed, in all but one of
the tests here, velocity decreased with crimping. The reason will likely be that it delayed opening the case mouth with the result that when the mouth opened and released gases that normally bypass the bullet until it obturates the bore. This would raise gas pressure so more escapes around the bullet before the bullet gets to the throat.
That doesn't always happen.
This article shows almost 5% increase in velocity in a 300 Win Mag. I've looked at data from a number of sources, and a claim of 5.5% increase in velocity from crimping for a 45-70 is the largest published one I've spotted. I've seen 2% in 357 Mag, personally, but I haven't done a proper study of pressure and crimping to examine this in all its variations. Also, the 5.5% velocity increase was with a cast bullet. An uncrimped lubricated cast bullet is so easy to pull out it almost counts as zero bullet pull, and it isn't surprising a crimp could make a difference there.
The area under the pressure curve is proportional to average pressure, which, divided by bore cross-sectional area, is the average force applied to the base of the bullet during its trip down the bore. The only way to increase the area under the curve using the same amount of powder and the same bullet is to increase pressure. So if the crimp increases velocity, pressure will go up, too.
To know how much peak pressure goes up is trickier. It varies with powder burn rate and barrel length and the rest. I've never heard of a gun burst being blamed on hard crimps, but the pressure addition isn't necessarily insignificant. If you get a 5.5% increase in velocity from a fast powder that normally burns up completely in the gun, the peak pressure may have to go up over 30% to get there. This is because that 5.5% extra velocity represents 11% more area under the curve with no change in muzzle pressure, as the powder was already being burnt up and was making all the gas for muzzle pressure that it could before the crimp was applied. Plus, the width of the peak is narrower with a fast powder, so it has to rise higher than a slow powder's peak does to make an equal increase in area under the curve. With slower powders, in many cases, the powder isn't all burning up in the gun in normal conditions, so part of the increase in velocity in that 5.5% instance will be due to increased post-peak and muzzle pressure from the more complete burning of that slower powder, and the actual rise in peak pressure might be as low as 11%.
In QuickLOAD, I've been able to create an example where pressure from a 5.5% velocity increase went up 37% with Norma R1, about the fastest burning powder out there. This still falls short of surpassing a maximum proof load. However, while I could create that example in the software using a 44 Magnum and Winchester 231 powder, that doesn't mean a 5.5% velocity increase could actually be obtained by using a hard crimp with R1 in the first place. Indeed, I suspect it is mainly with slow powders that an increase occurs, as a fast powder doesn't need a hard crimp to reach its potential. Because it has a shorter pressure rise time, bullet inertia alone does more to keep the powder confined without the help of a crimp. This is also why fast powders tend to be more position-insensitive.
Bottom line, while some situations can be theoretically created where there is a potential pressure hazard, I suspect the odds of getting all the way up to the minimum proof pressure by heavy crimping are slim, and getting past maximum proof pressure will take some real work. So while the warning may have some theoretical validity, I'd need to see an example to believe it.