However, he told me it is because PhD students have very specialised knowledge in a particular topic that makes it very hard -even for supervisors- to comment on it in a detailed manner. Well I still can't believe that's true in general.
Well if the implication is that your work as a PhD so specialised that nobody can give you useful feedback, I don't believe it either. If you cannot communicate your work in a manner that someone knowledgeable in the general area can understand and provide feedback, then your research should be doomed to fail since you won't be able to publish.
What is more likely is that your supervisor is insufficiently "incentivised" to put the effort required into understanding your work to give you useful feedback. There are many shades of grey here too. Different supervisors approach supervision at different levels of abstraction. Some are hands-off details. Some are hands-on details. Some need little incentive. Some need lots.
(Similarly, if you cannot interest someone enough in your research for long to engage them for feedback, I think your research is, in the longer term, also in deep water since you'll need these communication skills when applying for funding grants, to gain citations for your papers, etc.)
In any case, if you're looking for feedback other than your supervisor or colleagues ...
I'm not sure about in your area, but in Computer Science, there are ample ways of getting feedback on your work at conferences and other such venues.
Typically these conferences hold events specifically for mentoring students called a "Doctoral Symposium", "Doctoral Consortium", "Mentoring Lunches", etc. The format is different for each conference, but typically students submit a paper outlining their topic; the paper is peer-reviewed under special criteria. If accepted, the student is paired with one or two "mentors" (senior researchers who know something about the topic) at the event. The student presents their work to the broader symposium and afterwards gets some alone time with their mentor(s) who are typically instructed to be friendly to the student, but also to give some tough love if needed. As a bonus, many conferences publish the papers from these events in their proceedings, giving you a publication.
Another excellent opportunity for feedback is to attend a relevant summer school. At least in some of the schools I've lectured at, in between talks, the mentors have provided ample time to talk with students about their topics and plans, where at the end of the week, each student goes away with the perspectives of three or four mentors as well as a multitude of their peers. (More interactive schools tend to be in remote locations; everyone hangs around afterwards. Less interactive schools tend to be in universities; everyone goes home afterwards.)
Also, think about getting your PhD topic accepted for a poster session somewhere. PhD posters can be a good way to get casual feedback from a wide range of folks.
If you don't want to go so far away, organise a talk in your school and invite people. Put a lot of effort into making the talk engaging. Try to invite as broad a range of folks as possible and try to get feedback from your supervisor on the talk itself beforehand. Present and take questions.
The simple catch with all these methods of feedback is to think: how can I make people want to give me feedback, how can I make people comfortable to give me negative feedback, etc. @Jeffiekins suggestion of beer and pizza is a good one. Generally you should never expect feedback; you should earn feedback.
Like in a poster session, if your poster looks unappealing and you look disinterested standing in front of it, and instead of engaging with the person, if you simply read through the text-heavy poster while they wait, then giving you feedback certainly won't be their first priority.
Similar principles apply for your advisor (and elsewhere).
