I'd try just plugging in the 6L6's and watching for signs of redplating. They may be fine as-is.
That said, 6L6's usually require a bit more bias voltage than EL34's, so the bias may need to be made more-negative to keep the tubes from overheating.
To do that, make R49 a little bit smaller or R48 a little bit bigger (using the part numbers found in the schematic you posted).
It's easier to make R49 smaller; to do that, wire a 500k or 1M pot across R49 and wire as a variable resistor. That is, one wire goes to the middle lug and another goes to one outside lug. These will be the leads of the pot which go on either end of R49. Before soldering in place, verify with a meter that the shaft is twisted so the pot looks like its full 1M or 500k resistance.
Be certain the pot and wires can't short to anything. Now power up the amp with the output tubes out of the amp. Measure the bias voltage at pin 5 of the output tubes. Adjust the pot for the desired amount of bias voltage. When done, shut of the amp and verify there is no voltage present in the bias circuit. Measure the resistance of R49 in parallel with your pot. Find the nearest standard resistor value to what you measured, and install it.
"Desired bias voltage" is either a number your predetermined, or the amount of bias that gave the desired idle current when you reinstalled the output tubes. I wanted you to measure voltage at pin 5 to be sure you didn't make a mistake that shorted out the bias voltage, which would cause the output tubes to redplate and melt.
The amp as it sits doesn't have adjustable bias. You might choose to make R49 smaller and replace R48 with a pot in series with a resistor to allow the bias voltage to be adjusted within some range. I don't know offhand what range would be best; the schematic says -38v is stock with EL34's, so I might guess that if your final circuit can be adjusted between -50v and -30v, that would work well for a range of both EL34's and 6L6's.
The larger bias voltage will, as you guessed, fight against the trem and give a less-full trem sound. I think there's another thread around here with experiments on boosting the trem depth of this exact amp.
EDIT: I see Alerich suggested largely the same while I was typing.