It's often worthwhile building buffered test points into a circuit.

My comment, however, was directed at the notion that the newbie would be overwhelmed with a FET buffer, not because of its complexity, given that your initial reference to that NSC app-note, but because of the other remark I made, namely that one routinely expects an oscillator circuit, wired as specified in the various datasheets and app-notes and using specified components, to oscillate. One doesn't, therefore, build in test point buffers for that particular part of the circuit, even if it later turns out to have been advantageous. That is a part of the circuit that one has every right to expect to work correctly 100% of the time, and in 100% of cases. Even one failure, of the billions of parts that they ship, would give me pause. It should not be possible for such a vital circuit to fail. Unfortunately, it is.

I ask, however, Do we know enough about this particular situation to make any suggestion in repairing it?

BTW, I just fooled around for a few moments with 1x probes on a frequency counter in parallel with my oscilloscope, on a CMOS oscillator, and, what I found was that the oscillator was distorted, attenuated, but the thing kept right on running and drove both the counter and the oscilloscope. That was at very nearly 20 MHz.

Do you suppose something's broken in the O/P's circuit?

RE