I have a 50 watt b-18 head and a couple of B-15's. The 50 watt head has significantly more usable volume than either the 25 watt cathode bias '64 or the fixed bias '66. Or the '71 B15N for that matter. The 50 watter really shines driving 2 B15 cabs. The difference has to be in the OT, cause there's no real reason to think a pair of 7027a would be louder than 6L6's.
I tried a B25 thinking it would be like the B-18 head, but very different: farts out way early -- well mine did. I had more usable volume with a B15.
I know what you mean, there is a volume difference. I've found that at loud levels, that extra volume didn't make that much of a difference.
I find that different speakers will also make a difference in how much you can get out of an amp. You can take the same amp with two different cabs and one can cut though better or have a nicer tone. The capacitors in the power supply will make a difference in usable volume and headroom. If the caps are performing below spec the usable volume will be lower. Just saying that it is a factor. If the amp is optimally biased, it also makes a difference in what you can get out of it. I guess I'm saying that a lot of factors come into play when it comes to an amp delivering an optimal musical output.
There is a difference between the 7027A and the 6L6GC. The 7027A can be run at higher plate voltage levels (600V vs 500V of the 6L6GC) and it can dissipate an extra 5 watts of heat on the plate. This means that you can run the tube at a higher voltage which allows you to squeeze more output wattage from the tube.
This is exactly what Ampeg did in their 7027A based amps. The plate voltage of the 68 B-15N with 6L6GC's is 450V. This is in contrast to 7027A based amps like the B18N (and B-15ND) that had a plate voltage of 505V, the B15S was run at 575V. As you mentioned, the output transformer used also makes a difference. The 50W 7027A based amps had a different transformer than the 30W amps. I don't know if it was simply a beefier transformer to take the higher wattage or if there was an impedance difference that helped them get more power as well. I don't have the specs on the one that they used. The point is, the higher voltage allowed them to bias the power tubes into a region that allowed more power out of the tubes.
It is true that some manufacturers were selling 7027A tubes labeled as 6L6GC and these were interchangeable. If you take an amp designed for the 7027A spec and install a 6L6GC, you might be pushing the tube. You can sometimes get away with doing this because many of the NOS tubes could be pushed beyond their published data sheet limits. The data sheet specs were intended to provide optimal service life. Some manufacturers like Fender would design amps with voltages higher than the published specs. Back then, tubes were of good quality, they were cheap and the designers knew that they could take the extra stress so they took advantage of this.