Another loco stalling on points issue

I think you will find the flanges are not contacting the bottom of the crossing moulding, Peco are pretty good at getting their mouldings right. (It takes a very coarse flange such as the old 'pizza cutter' flanges on Lima to foul on the bottom of the moulding.)

You have actually begun to identify the problem by seeing the 'balance' issue. The problem is that the point base has a slight upward bow imparted by the curved rails attempting to straighten - that's an issue Peco haven't corrected. What happens as a result is that the centre coupled wheel is raised slightly on the plastic crossing, and the loco balances with both of the coupled wheels either side out of contact with the rail - and there is no pick up!

If you take one of these points off the layout and sight along the rail top - which should be flat, i.e. horizontal - you will be able to see the upward bow. Comparison to a steel rule as a straightedge will make the slight deviation apparent. It's not gross, but it only needs a tiny airgap between tyre and rail to prevent current collection.

Cure. My preference is to glue down the point on the ply base.I have some steel blocks which sit on top for at least 24 hours as the glue goes off. (I do this on Peco live crossing points too, rail top should be horizontal, on all except intentional gradient transitions.)

Then again with the steel rule straightedge. If there is any hint of a rock across the crossing, use some fine abrasive paper on a small block to bring the plastic crossing down to match the metal rail top. Gently does it, there shouldn't be much to remove and avoid scratching the rail top.

This is typical 'learn as you go'.

Yes, live crossing is superior: 'electrofrog' in old Peco speak, but now we have 'unifrog' from Peco as their new standard, which can be wired for live or dead crossing. (Peco have to offer both options as a great many layouts are still constructed on the basis of using the points for power isolation, a bodge that simply won't go away...)

Suzie is quite right about the potential for DCC shorts on dead crossing points; but you haven't mentioned your DCC system tripping when a loco stalls on a point so that presumably isn't happening at present? Some wheel forms and 'arrangements' found in RTR OO are worse than others for this problem, I can definitely point at Hornby's class 30 and the flangeless wheels in their pacific trailing trucks among current products; there are probably more. (It is possible to modify these dead crossing points to eliminate the problem completely, essentially altering them to live-crossing-alike' but that is for another day...)

Update 2 - I have followed the good advice above, and worked through several possibilities as to why I'm still seeing the locos stalling. I've made sure the pick ups are properly in contact with all the wheels in relation to axle side to side movement (thanks Butler-Henderson), and while I was at it, I cleaned the treads, flanges etc to ensure good electrical contact. I'm still not sure if I'm getting a momentary short somehow but the DCC controller does not trip, and I checked to ensure the wheel treads are not shorting at the converging of rails at the frog. Overall, the situation has improved with complete stops being reduced, but I'm still seeing the locos regularly 'stuttering' through the same points each time (although not all interestingly). However when the locos do completely stop it inevitably is when the central coupled wheel is over the plastic of the frog. When this happens, to get the loco moving again it either needs a slight downward pressure on the cab, or funnel, or slight sideways movement across the wheel axles. Of course, there is no such problem if the loco speed is increased.
I'm coming to the conclusion that I need to replace one of the points for an elecrofrog with associated IRJ to prove this cures the issue once and for all, then replace all the others likewise.

If you are still getting stuttering and the command station stays stable (ie no cut-out due to short), it would suggest to me that there is a loss of power.
Although you've checked that the pickups are contacting the backs of the wheels, it doesn't necessarily mean that they are making good electrical contact.
If you haven't done so already, may I suggest that you position you loco upside down and connect two wires from the rails to the wheels on one wheel set at a time while power is running. The purpose of this is to check the continuity of each individual wheelset and its pickups. Wobble the wheels a little to see of the motor stops - I highly suspect that you may have some axles which are not fully making contact, in which case, the pickups may need tightening up a bit more.

It could also be an issue of lifting. Since centre driving wheels on a crossing seem to be a common factor, it suggests that there might not be any vertical movement possible in the other wheelsets, although this would be unusual. Having said that, I have filed groves in chassis base plates before to enable a little axle vertical movement for exactly these situations. Be careful with this approach as it can be a case of solving the symptoms (stuttering, which some people mis-identify and end up using stay-alive) rather than the root problem - track which actually is raised - try my mirror technique above to see if the rails are raised.

If that doesn't cure the issue, then yes, electrofrog is the way to go.

HTH