Any recommendations as to what continuous current mins. for a 75A motor? I have seen some places where the recommend twice or more of the motor's current rating (for me 150A or more) and other saying that motors usually don't get to the rated current and you only need like 10% more (82.5A). Which is more accurate to what is needed to safely run the motor?
First off, for a motor this size, I wouldn't even bother trying to build an h-bridge to control it, unless I was an expert on h-bridge design (even then - probably not). If you're not, you'll likely end up spending more money blowing expensive mosfets than you would have had you simply purchased the controller outright.
You are mainly interested in one rating for the motor, and that's the stall rating. Stall occurs when the motor is idle (not turning) - when you apply voltage, the motor is technically stalled, so your motor controller will see (for an instant) the stall current. If this is higher than what the motor controller can supply, either the motor won't turn (because the controller will be in a fault condition), or the controller will burn out or be severely degraded over time (depends on how much you spent on your controller).
To measure the stall current on such a motor (assuming you can't find the specs from the manufacturer - try to look there, call them up, etc - first; it's just easier), remove the motor from the gearbox (you may need to find out how to do this first; some are easy, some are a PITA), then hook a multimeter across the terminals and set it for resistance reading. Slowly rotate the motor shaft, and take several coil measurements. Once you have your measurement, average them. That is the average coil resistance. Apply Ohm's law for the voltage you plan to run the motor at to get the current. That will be the approximate stall current. Add 15 to 20% more amps just to be safe.
Now - once you know that, then you can shop for a motor controller. Let me tell you this, though: For a motor that size, it won't be cheap.
Vantec - VANTEC Vantec RDFR, RSFR, RET, CDFR families
Roboteq - http://roboteq.com/brushed-dc-motor-controllers/nxtgen-controllers-selector
Again - expect to spend several hundred on such a controller for your size motors. Also again - that is cheap; you likely won't be able to build such a thing without extreme experience.
One last option - if you insist on building your own controller:
Get yourself a DC-controlled DPDT contactor rated for the current needing to be switched.
Connect the contactor in standard h-bridge fashion to control the motor direction.
Get yourself some beefy logic-level n-channel mosfets; hook in parallel, double the current rating you need for the motor.
Hook the mosfet bank on the low-side of the h-bridge (between it and ground). Hook it to a heatsink for insurance.
Use the h-bridge contactor to control the direction of the motor, use PWM on the to control the mosfet bank to control motor speed.
Be sure to set the PWM to 0 before switching direction of the h-bridge (unless you want your contact welded or worse).
Now note - I've left out more than a bit on the above, but that's the gist to make a cheap high-powered h-bridge that won't kill you in the wallet over time; the problem with trying to build your own h-bridge is that finding cheap and matching p-channel mosfets isn't easy, so most go for an all n-channel design; the problem there is, of course, the high-side drivers. So, you need some kind of mosfet bridge driver chip that can handle the everything your h-bridge needs, etc - and hope you don't blow much up along the way, wasting money, etc.
Honestly, if you want to do this right - you'll just take your motor measurements and purchase an h-bridge (unless you plan on learning h-bridge design and money isn't an object - if so and if not - go for it).