Whether you use a strain gauge or rely on pressure signs on the brass to judge the safe working limits, ultimately the choice of pressing the trigger on the next load increment comes down to you the shooter/handloader. If the strain gauge indicates 55,000 PSI is it prudent to add another grain of powder if you are working on the assumption that 60K is the safe upper limit? Do the pressure indicators of previous rounds support going forward, or is it time to stop? Perhaps the days of reading the brass for signs of pressure are past us. Armed as we are with chronographs and computer programs, perhaps those of us who find is necessary to define the safe limits of our equipment, have come to the time where the use of technology to define those limits is mandatory. If I can afford to purchase powder by the keg, bullets by the 1000, and Euro brass when I'm too lazy to uniform North American brass, spending $150 on QuickLoad doesn't seem so terribly outrageous compared to replacing a bolt, an action, or an eyeball!
Life was simpler when all I did was load hunting grade ammo, accepting whatever accuracy good loading practices and the available components would produce, and accepting the velocity predictions of the loading manuals at face value. But when accuracy becomes the overriding consideration, it also becomes necessary to find the safe limits of pressure to ensure that when making small changes to a load it always works towards lower pressure, instead of saying, "Damn, I wonder if I could pull in that flier if I went up just another half grain." From the top the only direction to move is down.
The lesson I got from this experience is that my method of determining maximum pressure is no longer viable and has to change. As time passes, there will be other new propellants I will try, and they too might exhibit different characteristics from what I have come to expect. So I will run "new to me" loads on QuickLoad to ensure I'm not outside the parameters of safety and I will have my caliper in my shooting bag to ensure that real world shooting agrees with the computer prediction.