Surfer's academy teaches a specific way of producing content, which is to analyse what currently ranks, extract the common characteristics and write something that matches them. As a process it is coherent and the training explains it clearly. Content structure, heading hierarchy, covering the subtopics a reader expects and doing keyword research that reflects real intent are all taught competently, and someone with no background will come away with a workable process. My objection is conceptual and it runs through the whole thing.
The score is derived from correlation with pages that currently rank, and the training treats it as though it were an optimisation target. It is not. It is a description of what the current top results have in common, which includes the things that made them rank and a great deal of coincidence. Writing to match that pattern produces a page that resembles the existing results, which is not obviously a route to outranking them, and the material never seriously engages with the difference.
In places it does say the score is one input among several. In practice, everything about the tool and the training points the writer at the number. A trainee who works through this comes away believing that a higher score means better search performance, and that belief is the thing most likely to waste their time. The consequence shows up in the writing.
Content produced to hit a score has a recognisable shape, meaning the same subheadings, the same coverage, the same length, and the same absence of anything the competitors did not think of. It is optimised into sameness. The material has no answer to this because the answer would undermine the product, and the honest guidance would be that the score tells you the table stakes and everything that actually differentiates you sits outside it. The AI writing integration makes this worse rather than better.
Generating content and then optimising it towards a similarity score produces material that is derivative twice over, and the training treats this as an efficiency gain. For a publisher trying to build any kind of distinctive presence it is the opposite. The larger omission is everything else about search. Links, site authority, technical health, brand signals, how a page earns citation in AI generated answers and whether anyone actually wants the page you are writing, none of that is in scope.
The training presents on page optimisation as the discipline when it is one part of it, and someone whose entire education is this will be confidently wrong about why their content is not ranking. Three point three. Clear, well structured, free training on a real process, marked down because it teaches a correlation score as a target and produces writers who optimise towards the average rather than towards a reader.