That would get you closer, yes.  How close?  I dunno without the math.  But, it's not going to hurt the amp to run it that way long term.

You could also try a NOS 5V4 rectifier, which falls somewhere inbetween 5Y3 and 5AR4 in sag and voltage, and is still a 2A tube.  I bet a 5V4 and 5881 would dial in pretty close to "just right."

Personally, if it were my amp, I would take the measurements with both the 5AR4 and 5Y3, do the math and figure out the correct cathode-bias resistor value.  I expect it would be either a 250-ohm or a 220-ohm resistor.  BUT if you don't feel confident checking voltages, I don't want to encourage you to do so, not without someone with more experience there to observe. 

Nik lists his voltage reading to ground in a chart at the bottom of his links page, but for biasing you need to "actual plate voltage" (voltage measured between pins 3 and 8 on a 6L6GC or 5881) and the voltage across the cathode-bias resistor, then it's a matter of plugin them in here in the "Calculate Plate Dissipation Based On Plate Voltage And Cathode Current Readings" section.

http://www.webervst.com/tubes1/calcbias.htm

First:
# of tubes sharing the resistor: 2
voltage across resistor with 5AR4: ??
actual resistor value: 270
actually plate voltage with 5AR4: ??

See what the calculated current and dissipation is.

The try it with the 5Y3:
# of tubes sharing the resistor: 2
voltage across resistor with 5Y3: ??
actual resistor value (start with stock): 270
actually plate voltage with 5Y3: ??

The start trying lower resistor values until you get close, without exceeding the plate dissipation you found above.