"if people were better educated, yes, anybody could gain the benefits of the healing effect" -- that's a fairly optimistic view, not born out by evidence. In fact, there's significant variability in the individual sensitivity to the placebo effect, and evidence that this variability is driven by genetics.

Here's a relevant quote from "The placebo response in human evolution" Medical Hypotheses (2005) 64, 414–416
"Being the ‘‘enhanced CNS–body interactions/placebo sensitivity’’ a biological trait, it was and is susceptible to bear genetic polymorphisms and suffer spontaneous mutations; this would explain the well known interindividual variation in the response to placebo [27]"

The bottom line is that some individuals may be responding to placebos like homeopathy not because of their "mindset" and "ignorance of science", but because they are genetically wired that way

As far as the potential harm to science and "real" medicine, I find that such concerns are overly paranoid. If a homeopathic medicine doesn't treat someone's pain, they will likely switch to a synthetic substance, so in the end, the best pain-killer wins.

Randi never says that it harms science/medicine, his objection is that homeopathic medicine is based on deception. However, I contend that it's "benign deception", and may result in an improved quality of life to people with proper genes.

The real debate is not whether people healing themselves through self-deception harms science, but what the doctors should do about it.

Consider the two scenarios:

1. An old lady comes to the doctor's office asking to renew her prescription for a tonic, which she says did wonders for her back. The doctor knows the tonic has no active ingredients, should he renew?

2. The doctor has recently learned that fake brain surgery that which actually just drills holes in the skull but does no brain manipulation has a significant benefit to Parkinson sufferers. However for the benefit to occur, the person has to think that real surgery took place. Should the doctor recommend this procedure? Should they charge the cost of full surgery in order to enhance the "expectation effect"?

Both examples are real situations faced by doctors I came across when doing some research on this. The first case seems to be pretty clear cut -- tonics have no harmful effects, and are cheap to boot, so most doctors would renew. The second case is much more controversial, especially if they charge for the full procedure.