Well, in theory if the spindle has repeatable runout, perhaps. In practice, you'll be sparking out for an hour. I say that as a person that has recently rebuilt his surface grinder spindle and instead of 10 minute spark out, I now do 2 passes. That's the difference runout (especially random runout) makes.
Still, this never has stopped me from using tiny inadequate spindles for precision grinding. Mostly for cylindrical grinding thought as I don't have a proper machine for it. I use a proxxon-made dremel like spindle. I never measured it's runout, but it is made for drilling PCBs and it can do it for hours without breaking 0.4mm (16 thou) carbide drill bits so the runout is pretty low(compared with dremels, in surface grinding world it is still huge). If I set it up right I can get really nice finishes and acceptable accuracy.
But recently I was forced to use it with a really long stick out of the little abrasive wheel. This was for ID taper grinding. I lowered the spindle rpm and I upped my workpiece rpm to still get the correct surface speed, but the finish definitely suffered a lot. You can see an example here:

There are concentric scratches. They can't be felt with a finger, and this is a functional part so it is fine, but it could be unacceptable for something visible.
On a surface grinder, depending on table feed speed you'll get waves, repeating scratches, stairstep patterns.... You can chase it all day long with a less than great spindle... There are ways to work around it (randomising your table moves, and lots of time). And there is always lapping.
Personally I'd rather use a surface grinder as a T&C grinder than the other way around.