One thing you don't mention, and I'm quite confused about, is what your PhD would actually be in. Web development and php, while important and valuable skills, do not prepare you for any sort of research I can think of.
On top of this, 6+ years is a long time to have been out of a field. Natural language processing, visual info processing and distributed programming have moved on tremendously in that time. Unless you've been doing something day to day which keeps you up to date with the cutting edge, you're going to be very behind.
On the other hand...
Research isn't about getting grades - it's about being able to publish work that is useful to others. When you applied before, all people had to judge your potential to achieve this was your (average) grades. Now though, you've been out in the wild for a few years doing stuff. If you can present things you've done which will convince a potential supervisor that you can come up with novel solutions of a quality high enough to get into a journal, then you've got a shot at it. My (wild) guess would be you should aim for a lab that does applied stuff rather than theoretical stuff, where your prior day to day coding experience might shine through.
As an aside, have you considered applying to get chartered status in your field rather than a doctorate? It's challenging to get so generally highly thought of (this varies country to country and profession to profession though) and more closely based on industrial experience. If you have six years of that under your belt, along with a relevant masters degree, then you may be well on the way already. The yearly fees once you have it can be expensive, but worth checking out if you haven't already.