A perfect sine would give you no harmonics.

I'm pretty sure that the harmonics is what gives a sine out of an
oscillator it's character. The ASys sine is not bad, that's a
personal opinion. The waveforms on my scope show the M15 as the most
mathematically sine like, the ASys has slightly pointier peaks (and
thus sounds brighter towards a triangle sound), and the Z3000 has the
most "imperfect" waveform of the three mentioned. However, the Z3000
and M15 FM better than the ASys (which has no dedicated FM input).
The ASys however has a wave shape control that lets you skew waveforms
and morph saws from ramping down to ramping up, and goes into LFO much
slower than the M15.

There are no "bad" tools, just lack of understanding and/or the wrong
tool for that particular job.
I use the scope to see what modulations do to signals, not to judge if
my ears are doing their job. I hear with my ears, not my eyes.

If mathematical perfection is that important, perhaps CSound is more
what you want.