So here is the presentation section of my personal studio. I will try to explain what motivated the choice of the machines that can be found there.
You will find that there is a very different set of machines, some very old, some very recent.
I concive the synthesizer as a real instrument. I therefore accept the fact that it has a certain sound color of its own, so either we will like this sound color or indeed not. Some machines, such as the Akai S-5000 or S-2000 samplers, have a very transparent sound colorization, that is, they do not add or remove almost anything at the initial sound.
On the opposite, the EMU Emax S.E. is very different and has a natural predisposition to add a metallic and organic behavior to the sound (a slightly rough / granular sound if you want).
One can also make the comparison with guitars. Some guitars will be very welcome to make metal rock or very heavy, and behave badly enough to play funk for instance. Conversely some guitars are perfect for funk or jazz but will be poorly adapted for other kind of musics.
The various major manufacturers have clearly understood this and we find a typical sound related to a chosen brand. An Oberheim will sound very different from a Roland (similar technology goes without saying). Even if the filter and oscillator settings are exactly the same.
Indeed, for instance, a primitive sawtooth wave coming from a synth will have a steeper slope with a particular amplitude and a phase angle shift compared to the same sawtooth wave coming from another synth. So, the characteristics of the amplification (if we speak about analog technology) or a D/A conversion (the Digital to Analog converters, which will be found in the digital synthesizers) will greatly influence the way the sound will be produced.
In my example of a sawtooth wave produced by a synthesizer X (let’s say a microKorg) and a synthesizer Y (for instance a SH-201), we can hear very clearly the two sounds produced have great differences on the spectrum frequencies analysis. Some programs, like Steinberg Nuendo, are very helpful to get a very elaborate frequency graph, allowing us to visualize the sound variations in an extremely precise way !

It seems to me appropriate to have a bunch of machines from different manufacturers to be able to use a specific equipment for each category of sound. That’s why I’ve chosen to pick some machines from different manufacturers.
The other essential topic is the rendering engine of those machines. In other words, the synthesis orientation and the sound production process (analog circuits, digital DSP, …).
Here too I tried to make my choices on machines that allow me to have a good balance between all the different technologies. So I’ll use the Juno6 sounds for some things, or the Alpha Juno 1 for others, or the Ultranova … It all depends on the type of sound that you want to create, and therefor the synthesis method and the appropriate rendering process.
Obviously about the « Virtual Analog » synthesizers, one may wonder if it is necessary to use a true analog sound or not. Especially when these sounds go through a set of effects (chorus, phaser, equalizer, delay, reverb, …) some subtleties will be either erased or amplified…

There is no rational answer to this question … It’s above all a matter of feeling. For instance, the Yamaha AN1X (V.A. synthesis and therefore a 100% digital process) creates pads with a warmth and a texture much more gentle than a real analog synthesizer, like for instance, the Roland Juno-6, which may seem incredibly paradoxical. But it must be borne in mind the fact that a synth is a true analog does not guarantee in any way that a sound will be warm or « rounded ».
Some analog synths will sound cooler than others, using however a digital engine …
And this also applies to VA type synthesizers where some algorithms allow a warm and soft rendering of a sound, while other manufacturers propose algorithms rather badly adapted to produce warm sounds, but very efficient to produce slamming or aggressive sounds…
So with a same technology (analog/digital/va) we may get completly different results !