Try a voltage divider for base bias. This may give better results than the "suicide bias" shown.
May also want to use lower resistors, so the output current is a bit higher. That'll limit performance if the load is heavy. Likewise using larger capacitors (say 1 or 10 uF electrolytic, positive terminal towards the circuit) will maintain low frequency response when that is the case.
I would suggest:
R2 = 10k
R1 = 220k
Add resistor from base to GND, 33k.
This should amplify cleanly, and shouldn't be disturbed much by reading with a voltmeter. You shouldn't see much if any distortion, at least until it's really kicking (>3Vpk output voltage swing). Once it's verified working, you can change the added resistor and see how that affects the sound.
2N4401 is fine, 2N4001 I assume was just a typo. Mind that if you got them from unverified sellers, you might have, who knows what, floor sweepings, instead of actual transistors in there. If they're substituted with general purpose Chinese types, it should still work, this is a very noncritical circuit; the problem really is if whether they work at all.
May be worth setting up some basic test circuits to determine what the transistor is, BJT or MOS, N or P, EBC or what. A voltmeter and resistors should be fine for this.
Tim