The most glaring of which is the long and indeterminate start time of the crystal oscillator, if it will start at all. One reason the RC reset circuit normally illustrated with an 805x of the generation of yours, has so long a time constant, is that it must be long enough for the oscillator to start. The one you showed in your schematic will stop immediately, but it will take many, and that number is quite variable under the best of conditions, milliseconds to start.

That's why I wrote all that stuff about gating the oscillator on and off using a latch circuit.

That JMP $ won't help you if the oscillator hasn't settled into a steady state yet. There's no way to predict the behavior of your MCU under those conditions.

RE