Well, I don't recall having said that a 5% difference will definitely matter, so I'm confused as to why you're confused.
If I said as much, I must've "mistyped" (it's election season here -- no candidate ever makes a mistake now, they only "misspeak").
Here's what you need to make it sync:
1) The right polarities for sync and video
2) "Close enough" sweep rates (BOTH vertical AND horizontal)
What's "close enough?" There's no hard and fast rule. There are no guarantees when operating equipment outside of original design targets. Monitors differ in their tolerance of off-spec signals. That said, if you're within 5%, that's usually not a problem. In some cases, even 10-20% might work. As you go beyond about 15%, the population of monitors that will sync drops off rapidly.
I have plenty of experience interfacing compact macs to TTL monitors of the type used in the early IBM PCs. Almost none of those sync natively; they generally require internal mods to their horizontal sweep circuits to work with the Mac sweep rates. The vertical sweep presents no problem that I recall.
The necessary mods are easy to implement, but difficult to deduce without a schematic. In many monitors of that era, the horizontal sync circuits were implemented with the venerable 555 timer chip. The sweep rates were set by precision resistors and capacitors (typically of 1% and 5% tolerance, respectively). So, if you're able to identify the horizontal circuitry, and additionally can zero in on the timing components, you can rapidly calculate the value of resistor to put in parallel with the existing one to push the frequency upward by the necessary amount.
If you want to use a potentiometer instead of a fixed resistor (not recommended), be sure to put a resistance in series with it, to prevent ever going to ~0 ohms at one extreme of the pot's adjustment. The series resistance should be chosen to cause the sweep frequency to exceed the target by no more than 10-15% when the pot is at its minimum resistance.