Using piling (drive pipe) to restrain water pipe

Have not seen that technique although it may work. It is known as a batter pile.

It would not seem to be cost effective as compared to mechanical joint connections. It would also be difficult to install the driven pipe precisely where it is required.

Dipra mentions this technique when repairs are necessary for unrestrained pipe:

"These precautions might include supplementary thrust blocking, restraint with laterally loaded piles or batter piles at the thrust focus, special pipe anchors, or other careful, sequential, and innovative engineering and construction procedures."

Yes, in a sense (to add to what bimr has said), actually as the soil is really what large thrust loads are/must be transferred to with many systems of thrust restraint, any time one has to remove the soil from around a working line that can't be shut down for any reason, any low footprint but well thought out methods could be helpful.
I saw a presentation a few decades ago from a gentleman I believe promoting something like this, it appeared for restraining fittings at that time I think with application to asbestos cement piping (that really didn't have an option of effective restrained joints). Reynold Watkins et al per must have at some point seen the like, that I remembered were then called "thrust pins". It appeared an advantage to this method e.g. was relatively easy constructability e.g. driving a pile or boring a hole with commonly available augers, though I suspect disadvantages include that someone likely has to design and place vertical etc. structural reinforcement for any very great loads, and also unless the contact between the pin and the bend has somehow pretty good bearing area there could be questions of contact stress and/or also unplanned lateral "rocking" of the bend due to a little misalignment or hydrodynamic effects (one would not worry about with a properly designed and poured concrete, wide buttress blocking).
That being said I actually happened to see sort of a cross between the two employed also a few decades ago to restrain a 48" bend in a ductile iron piping system for an ~300 psi field test, with really no competent soils directly behind same. What I believe the designer did in that case was dug a quite deep hole somehow directly behind the fitting and then at much shallower depth leveled a relatively shallow bracing pad or skirt that fanned out from that hole in the shape of a normal block, set some substantial reinforcement vertically into the deep hole with lesser horizontal reinforcement tying out into the skirt (the whole thing in profile view behind the bend looking like an upside down "L"), poured the whole thing up to not much far above the pipe with concrete, and finally embedded a bolt into the top of the block that was actually exposed during the test. I was told later that with surveying instruments trained on the bolt during the subsequently reported successful test, there was "no movement" when that more than a million pounds of thrust were applied at that location (of a sort of cantilever-loaded vertical pier with stabilizing cap).