For normal use, you might want to consider a latex/acrylic bonding agent with the proposed cleaning; I've used them with good success.  I've had a mix of success with just wetting.

The forklift traffic can be a little spooky and may cause some heartburn; it's a combination of the high wheel loads, hard wheels, speed, and turning.  They're pretty rough on a bonded overlay.  Scaling is possible with anything short of an epoxy bonding agent (very costly).

Can the adjustment be taken by whatever racking system is proposed or do you really need a flat floor?  Is the 3/8 rock a 'trap' rock hardener?

The 15 or 25 psf in dead load shouldn't be a problem with the strength of the slab, but the change in 'real' dead load may have an impact on the supporting structure and might cause some differential movement.

I agree with dik...you would be surprised at the relatively high stresses at the bond interface with the forklift traffic on a thin overlay.  Evaluate with elastic layer analysis to see magnitude.

You might also want to try a bit of innovation by using some curling restraint near the perimeter.  This can be done by installing small (1/8-inch) threaded rod "dowel loops" vertically in the existing concrete prior to placing the topping.  I would put these in two rows, staggered, 12 to 15 inches apart at the perimeter.  This will help attenuate the curling stress affect on the vertical bond.  The loops will look like an upside-down "U", installed in drilled holes in the concrete and set with epoxy.  The "spread" on the "U" should be at least 4 inches and should extend to about 1/2 of the overlay thickness.  Place them perpendicular to the expected directio of maximum shrinkage so that they don't provide excessive restraint to the shrinkage.

Both, very helpful input.  Thank you.  Do either of you feel that bond testing would be useful.  An if so, what would be a criteria on which to base testing success or failure?

dik,  I didn't understand the 'trap' rock?

The floor flatness is not an issue... at least not one that I am concerned with on this project, due to my confidence in this particular contractor.

Ron,  The system that was used along the perimeter of the first pour (84' x 84' = 7056 sf) was a [looked like] 12 ga mesh in a 2" x 2" pattern, about 2-feet wide, centered on the construction joint and held off the structural slab with (2) #4 bars as chairs.  This was a curling-restraint idea that the construction manager had done on a previous job with reportedly good success.  We'll see....  

Thanks again for your thoughts.

I've never tested latex or acrylic bonding agents; there are ASTM tests that the materials have to meet.  Anything that 'sticky' to apply has to hold something <G>.

The comment regarding the degree of flatness required for racking, etc. was to rethink if a bonded topping is required.

'Trap' rock is a dense, fine grained igneous material that crushed into small particles  and is broadcast (along with additional cement powder; typical surface hardener) and embedded in the surface of a concrete slab to provide a hard durable wearing surface.

I've used 4x4 6/6 WWM along with #6 pins drilled into the concrete and gently tapped in place with a sledge hammer (mechanical fit only) in a fashion similar to Ron's and your construction manager; these took the place of a support bar and the mesh was tied to the pins.  Although not used, it's possible to use powder actuated fasteners with 'holes' in the end to wire the mesh to.  This provided mechanical connector/reinforcing at edges.

Although it's located at the bottom of the topping, I'd be reluctant to have a 1/2" continuous bar for support.  The elastically 'hard' material could create a line of weakness (As Ron noted, try a FE model and see what tensile stressed are created as a result of the bar).

JSA...with regard to bond testing, use the ACI method (2" diameter core, direct tensile pull).  The criterion for acceptance is tensile failure in the material it is bonded to (original concrete), without significant regard to the quantification of that bond strength.

Ron:
Thanks... wasn't aware that the ACI had a method.

Gentlemen, thanks again.

At the risk of overstaying my welcome, both of you value the results of an FE analysis as useful information for this problem.  I wouldn't have assumed I could get useful info from one.  Could either of you explain to me how I would model this to get the information which you feel would be helpful?  I'm imagining that you would need to quantify a shrinkage amount of the topping slab, shorten the topping, the shortening is resisted by the in-plane stiffness of the structural slab, which gains some in-plane stress and then.............. are you quantifying the bond stress somehow?

Without getting into a 'real' discussion (still stiff and sore from the last one), the FEM is a useful tool.

I wouldn't normally use it for this application, because I already know what happens and have an MS Word template 'disclaimer' to send to the client that includes an attached 'picture'. Somewhere you should be describing the loading and area of loading anticipated and the problems in using a bonded overlay.

It was mentioned in the earlier discussion just to provide additional info on what happens and what can easily be illustrated.  From a liability vantage, if you're into it, it is useful 'due diligence' when it delaminates and you haven't covered yourself with correspondence. You should have some record of a caution about the bonded topping not working for a variety of reasons; the next owner is going to be running hard wheeled forklifts all over and going to be lifting pallets of cement bags and is going to third party you when the floor fails.

The change in stiffness of the interface by the bonding agent increases both static stresses as well as dynamic stresses.  For the latter it offers a reflection surface as well as increases the amplitude of the shock wave. (FEM can illustrate that).  

Similarly, the hard rebar also increases tensile stresses when a load is applied (ditto re FEM).