It is not uncommon for OEM manufacturers of pumps, compressors, and blowers (variable torque loads) to supply motors with their equipment which must run at the service factor to provide the necessary power to the load. Personally, I disagree with the practice, but there is (somewhat?) of a justification for doing so. I am willing to go into that if anyone is curious.
In your case, the original rating of 32A at 230V is the power requirement of the pump. The motor rating of 28A at that voltage for a 10hp motor is correct by NEMA standards. The pump manufacturer's assertion that the motor can operate at 32A continuously in this application is correct if the motor has a service factor of 1.15 and the ambient temperature is 25C or less (ie 28*1.15=32). Beyond this your situation seems sketchy.
In your case, the namplate data for the new rating (212V) is somewhat ambiguous. Superficially it makes sense that with a reduction in voltage from 230 to 212 that the service factor will (should, could?) change from 1.15 to 1.28 to allow the maximum allowable current to increase proportional (from 32-36) to the reduced voltage. However, the idea that a proportional increase in current will offset a decrease in voltage with the resulting power being the same is a fallacy because of the relationship between volt/hertz and torque (power) in an induction motor.
It would appear from your post that the only nameplate values that were changed when your motor was re-rated were the voltage and service factor. If in fact you have motor whose nameplate states 10hp at: Xrpm, 212V, 32A, and 1.32SF and the motor current when supplying a 10hp load is 36A then I would question that rating based on NEMA and EASA standards for motor ratings and re-ratings respectively.
With respect to the heat exchanger fan direction, it makes sense that with a totally enclosed motor (specifically totally enclosed, not TEFC) on a "package" motor/load system that the heat exchanger fan would blow away fom the motor since a totally enclosed motor rating is based strictly on ambient temperature. This is because the only cooling provided (or necessary) for this frame type is internal to the motor. If the motor is exposed to greater than ambient temperatures (ie air blowing across it after passing though a heat exchanger) then it would have to be derated.
In sum, I would be very suspicious of the situation you are describing based on the information provided. In the best case you have a motor that was actually rewound to provide NEMA rated hp (and the implied service life) but has an incorrect nameplate. I doubt whether this is the case. In the worst case you have a motor that is not properly rated (or nameplated) and will last perhaps as long as the warranty for the equipment with which it was supplied. As a test of where you stand, take a temperature reading of the stator of the motor in question and compare it to the temperature rating of the insulation class.