The problem with dippers (and actually, any powder measure that isn't a scale) is you are trying to take a substance that needs to be measured by avoirdupois (actual weight) and measuring it by volume instead. Now, it is possible to convert a volume to equate to a certain weight. This is what Lee does with the VMI tables, assigns a certain number which is the way to convert actual grains to the CC of the dipper or measure cavity.
The problems arise when you think about the fact that powder changes its weight based on humidity or formulation changes lot-to-lot. A batch of powder made two years ago is certainly going to be different than a batch made today, and a humid day makes them different again.
To cover their @sses from liability and keep us from blowing things up, they "fudge" the tables on the worst-case-scenario side. If you get the most dense lot imaginable of a certain powder, the Dipper Size "X" cc's will not throw any more than Y grains based on the VMI chart.
So in most cases, dipper size "X" will actually be less than you want in the real world. However, if you do your part to make each dip of powder consistent and level off each charge the same way, they will throw very close to identical each time. What you need to do it practice and then measure what you are actually throwing, and use the size dipper that gives you what you want consistently. It might be the "X" cc dipper, it might be the one next to that "X" cc dipper.
The same goes for powder measures. The Lee Autodisc measures are very repeatable and easy to use but you are stuck with the cavities they provide. You need to find the one that matches up the best to what you want, regardless of the VMI table and what the Lee load data says.
If it is a choice between Lee's "We don't make our own data, but copy it from other sources" data (which they pick the LOWEST charges that fit the criteria, to prevent people from blowing stuff up) and the powder maker's data (Who uses actual test barrels and wants really good info available to reloaders) I would always take the powder makers' data.
That's not knocking Lee- they have a vested interest to provide as close to idiot-proof simple and cheap gear as possible and it works great for what it is.
Another thing is that from what you said earlier, Lee's data was for a jacketed 200-grain whereas the powder maker's data was specifically for a Speer Gold Dot Hollow Point. Different bullets have different bearing surfaces. Some will push through the barrel easier than others. That's why the specific bullet is mentioned to duplicate that load. Hodgdon, for example, tests a variety of the same style & size of bullet, and publishes the MOST restrictive load. That way, you are never exceeding that load with any other similar size and style bullet. So there's a built-in safety margin. What Lee does is find the most restrictive load data they can and then round it down to the dipper that will serve using their VMI table which already has a safety margin built in to compensate for density changes in the powder volume. You now have at least 3 safety margins stacking up to drop your charge down from nominal.
That's not necessarily a bad thing especially for a novice reloader, that way nobody blows something up.