Don't poke too hard as even well bonded glass may start to peel with enough pressure and a sharp putty knife.

This caught my eye when I read it.
I went down to the shop and pulled a deck section- hatch recess cutout - off the scrap pile and went to work with a putty knife.
I removed a small (4" x 4" ?) area of glass with the putty knife. About 80% of the removed glass had a thin (fraction of a mm) layer of wood attached- the wood fibers appear to have failed at the limit of epoxy 'penetration'.
It took a lot of effort and even some hammer tapping to get that separation, and then the glass fractured before I could remove a larger area. So that sample met my definition of 'adequate' - the wood sheared before the glass-wood bond.
So, it all depends on the definition of 'enough pressure' on the putty knife, I guess.

When I did have a problem with glass separation from wood that had been 'seal coated' , I was able to rip off a section of glass with my hands, once I got it started (at a blister) with a putty knife. I kept ripping the glass until the glass fractured, and feathered from those edges.

Interesting to me is the fact that that failed area was a black (epoxy) area on the cockpit floor, just in front of the seat - the area that gets full sun when the kayak is car-topped on a sunny day. (Yes, even in the PNW, it can get hot and sunny! :-) )
It's also an area that is usually wet when I'm paddling. So it may have been a situation where pinholes in the epoxy allowed some water into the wood layer, which caused the failure when later heated by the sun - the situation Rod Tait has described.

Luckily, wood-core epoxy fiberglass composite construction is so strong that for most of us, we never get close to its limits with regular paddling.