From your quote:

According to the Centers for Disease Control and Prevention, almost every major study on defensive gun use has found that Americans use their firearms defensively between 500,000 and 3 million times each year.

You didn't provide a link to the Heritage Foundation document you linked to, and I couldn't find anything from the CDC on the subject. It appears that what they're referring to was that the CDC used to say on their webpage that:

The stats sourced from a CDC-commissioned study finding that instances of defensive gun use occur between 60,000 and 2.5 million times per year. References to that study were deleted from the site following private meetings with gun control advocates in 2021, emails obtained and published by The Reload show.

foxnews.com - Cdc removed stats defensive gun use pressure gun control activists report

Note the difference between 60,000 on the low end of what the CDC actually said, and the 500K that the Heritage quote shows. That's probably the result of Heritage defining "major study" in such a way as to only include the very high estimates. I suspect they excluded the studies that use actual reported data to assess DGU rates, and instead only "counted" the ones that came in with a really high number.

That's not hard, because all of the very high-estimate "studies" are basically just surveys of people asking them (sometimes years after the fact) about whether they used their guns for DGU. But again, all of those "studies" are consistent only with themselves. They're completely, utterly, and absurdly inconsistent with any data for gun use that can actually be measured. You either end up with people owning guns experiencing crime at rates 10x or 100x that of other people within their same zip code, or people only reporting defensive gun use to the police 0.16% of the time (when those same surveys have people responding with a police report rate of 50% or higher).

The reason, of course, is that the survey method is utterly useless for figuring out the real rate at which DGU's occur. Because people misremember what happened, when it happened, or misattribute an instance where they "use" their gun on an innocent person as a DGU. As noted in the analysis that commonone cited above:

Kleck and Gertz often defend their paper by claiming that their results are consistent with the findings of other private surveys. They explain that the reliability of a survey should be judged by the degree to which it coheres with the estimates of other surveys. However, using a tool we know to be flawed, over and over again, does not increase the quality of estimates deriving from the tool'it merely produces convergence to an arbitrary number. Surveys, for example, regularly show that men have sex with women more often than women have sex with men. Survey results don't mean anything if they don't pass muster with reality.

politico.com - Defensive gun ownership myth

It's a myth, BHM. Those 2-3 million DGU figures are almost certainly off by a factor of close to 100x. Analysis of DGU that extrapolates from actual, measured data - rather than just people's recollections or survey responses - come in closer to 5K-30K.