Rats
I can't find the particular study now, and I've referred back to it probably 60 times in the past! I'll tell you what I remember now, but the danged study is sure to be online somewhere - I just keep getting dead links at the moment.
There was an agency in the NIH (headed by a G Reid Lyon who has since left the agency and works in research on learning disabilities) which for several years conducted good science-based research to tackle the question of reading failure in this country (about 35% of students fall under this category, I think). They examined everything, not just phonics, but a broad spectrum of issues that factor into reading skill. They didn't conduct just one study, but many, to assemble a picture of how students learn to read, and why they fail to learn to read. The research was longitudinal (following the students all the way through school). There was no pre-existing prejudice towards systematic phonics instruction at all, and it was really just one of many issues the research looked into. Their findings concluded that phonics instruction was just one necessary element--there are others as well, phonics alone isn't the whole ball of wax. And the research explored the nitty-gritty of which elements are necessary in the phonics instruction - what works and what doesn't.
As I said, they found that for about 30% of readers, phonics instruction doesn't really matter...these students can read well with no phonics at all. There wasn't a prejudice towards phonics, but the studies were intended to develop an understanding of reading failure, and that's where the phonics instruction emerged as 'non-negotiable', in their words. Students who struggle to learn to read
need phonics...that was the clear finding. They also found that there are about 2 students who show significant improvement with phonics instruction to every 1 where it doesn't make any difference.
Now for the caveats (sorry this is so long

). This research looked at students in schools. Tangled up in this phonics question is their findings that if a student who can't read well by age 9, the odds aren't good they will ever catch up and be successful in school. The odds are very much against them. If you read the analysis carefully, it becomes clear that much (but not all) of the reasons for this don't apply in the homeschool setting. A 9 year old who can't read well develops a bad self-image in comparison to peers, and in terms of the evaluations he or she receives from school, and students of course tend to live up to the expectations they have of themselves. Also school curriculum are overwhelmingly dependent upon textbooks...so a non-reader or struggling reader is at a disadvantage in nearly every subject. And a further reason is that a 9 year old has struggled in the typical school setting for several years already--several years of misery trying to read, building up a lot of stress toward the subject, which leads most to kind of 'shut down' and stop trying. These problems are hard to overcome in a typical school situation. They are easier to avoid in a homeschool situation.
So the point of this is it is a big deal if phonics instruction help students learn to read well faster in the context of public education. The research had demonstrated that reading well faster (at least by age 9) is practically essential. But as I explained, many of the reasons why early reading success is so critical there won't necessarily relate to homeschooled students.
The one issue that I can remember that would factor into homeschooled students too is that students who do develop reading struggles, those who you find who will
need phonics instruction to 'get it', are much easier to help
sooner rather than later. Even students who start off okay can then kind of plateau because they just haven't developed effective strategies on their own. They may rely on context or picture clues, and kind of 'good guessing' strategies based on the beginning of the word, for example, and that kind of skill doesn't build, or lend itself well to more complex reading. But if students rely on this in the beginning, it becomes a reinforced pattern that can be challenging to overcome. I've worked with a lot of struggling readers that have this problem. Again, these are the students who we know can't read without this help...that's not all students, but those that fall into the statistical 1/3 of students who fail to do well and really have to have the systematic phonics skill building. It is tough for them to break the bad habits of guessing first, and go with method second. And it is a bad habit because it interferes with good reading in other ways, especially in the fluidity and automaticity, which are ultimately critical elements as well to good reading skill.
This research aside, having worked with students with reading problems myself, if I had my way phonics would be cut loose from reading and firmly attached to spelling, and then composition, to start with. This is just maybe a crazy notion of mine, but I really think there's a lot to be said for this idea. We push reading too much early on, and that puts out the idea of 'right ways/wrong ways' to read a book to solve the problem the pushing created. There are excellent phonics curriculums out there that demonstrate English 'follows the rules' much better than most of us realize, and if students learn those rules to
spell, they can transfer the rules from composition to reading too.
Too much bad phonics instruction does hinder good reading. It destroys the lyrical 'sound' of good writing. And beginning readers do need to read
good writing, not just the 'mat had a fat cat'. I think Dr Suess was an excellent exception, but most early phonetic readers I've seen are ugly sounding blecht. Reading ultimately isn't a mechanical process, it's lyrical and literary one.
Linda