I know there are complicated electronic fluorescent starters for preheat fluorescent circuits that work just fine, but I am wondering specifically about the use of a simple SIDAC as a direct substitute for the traditional glow bottle starter. Here is what I am thinking:

With a SIDAC that has a carefully selected voltage, every half-cycle that the fluorescent tube doesn't strike will consist of a significant portion of the input waveform being devoted to heating the cathodes, and then an inductive flyback voltage once the SIDAC unlatches(?). This means that within 1/120th or 1/100th of a second (depending on if it is 60Hz or 50Hz) of the tube becoming warmed up enough to start, it will strike. I feel like this would be a very fast way to start a fluorescent tube. This in theory would eliminate any excessive heating of the cathodes that might contribute to increased wear.

Do you think this would work?

The SIDAC won't generate any inductive kick, because it switches off only by the current reaching virtually zero. So the only thing it will do is to warm up the electrosed.

What you need for the starter is first to be ON for a while without generating any spikes, you don't eant to ignite the lamp prematurely with cathodes not warm enough.
Only then you want it to start generating the ignition pulses until the lamp ignites.

So the component first should not and then should generate the pulses, you can not have both without some form of a control.

Such control is possible with specially designed thyristors: When ON, you may control the hold current ( the current, below which the thyristor can not hold itself and so switch off) by the negative gate current.
And give it a timer circuit, which will actvate such gate current after the determined preheat current (few diodes to create some voltage room, then transistor or other SCR and an RC timing cell; it then doubles as an EOL shut down, as after even more time it stops the thyristor even firing, so if the lamp is faulty, it stops stressing the ballast).
Plus of course another part of the circuit which is responsible to turn it ON when the lamp is not ignited yet (a resistor with a Zener).
And put a bridge rectifier in front of all of that, so it works for both polarities. And viola, you have a typical "advanced electronic starter"...

In fact, this was attempted as early 1960 as suitable components just started to appear. It is possible to generate some spikes with diacs that way, but hard to do that in a controllable manner. Most successful attempts to do so while keeping schematics complexity somewhat bearable were done by our British
friends, see here https://www.lighting-gallery.net/gallery/displayimage.php?album=2363&pos=76&pid=90214 http://www.lamptech.co.uk/Documents/Journals/Thorn%20Lighting%20Journal%2026.pdf and here https://www.lighting-gallery.net/gallery/displayimage.php?album=2363&pos=51&pid=120951

By the way, instead of just fantasizing, you can either run some simulations in SPICE (Multisim, LTspice, some open=source versions or models built-in in electronic CAD) or even play with actual inductors, semiconductors and oscilloscope. Doing so will yield you some very useful knowledge, there are many programmers around, but analog electronics designers are in great shortage...

Here is a collection of old schematic ideas all over the world from an old Soviet book involving mostly dinistors (unidirectional SIDAC style pnpn device) as fluorescent starters, some coming as early as 1959! Definitely what was later evolved into Vivatronics.

You also see there was a fun idea of timing electrode preheat by sensing the light generated by a discharge across the filament when it becomes hot enough. It evolved in a practical circuit that fixes that poor behavior of single tube 20W glow starter circuits, especially in home use.

Here is this Soviet starter device involoving a photoresistor and a couple of SCRs in a potted module. https://www.lighting-gallery.net/gallery/displayimage.php?album=7727&pos=5&pid=237688  http://www.155la3.ru/k232.htm

The advantage of the old T12 lamps was (or the "miniature" old T5 format like F4..13T5), they in fact do not need any inductive kick to start, the mains OCV is high enough. So then using the voltage triggered SCR (either as a component itself, or a generic thyristor with resistiove divider, capacitor and diac trigger circuit, which has an advantage of providing extra delay so allow for reignition spike of an ignited lamp to not trigger) in only one direction is indeed sufficient: One polarity heats up the cathodes when the lamp isn't lit yet (so the SCR gets triggered; the one polarity is also important so the DC component saturates the ballast to boost the preheat current), the other then unrestricted can attempt to ignite the lamp. Once the electrodes warm up enough and the lamp start, the ionization slow decay means the lamp reignites before the thyristor gets triggered.

But the higher ignition voltage of modern T8 tubes means this "sidac only" is not usable anymore. However maybe an addition of a FEC device may make the voltage to spike up high enough, but it then becomes more complex and expensive, so the special GTO thyristor based starters become not much more complex, yet still way superior performance, mainly because it offers decisive preheat with ignition really prevented so the lamp will never start not warmed enough, really very high ignition voltage in the 1kV range so it ignites even very stubborn Krypton filled T8 tubes, inherent EOL protection (the typical simple design makes it to stop triggering when the ignition is not successful for longer than few 100's ms; it goes this way as the turn off current for the thyristor gradually increases to find the best ignition point even with wide variation of the thyristor sensitivity, so much it prevents the circuit even from turning ON), so generally way more reliable setup.

And really getting familiar with the electronic simulators is a huge benefit,
and the generic lack of "old school" analog electronic (and also mechanical design) engineers (those that need to be able to really understand the physical part behind the things, whether it is electronic/electrical or mechanical or other similar) on the job market is just a fact of a job market world wide...

Alright, so I guess this will only work for tubes that would strike whe heated at mains voltage. Still useful, but not what I hoped.