Adding to pancho_hideboo's comments
The saw will not see any loss between it and the antenna therefore the return loss will vary by two times the insertion loss of the switch. You could say that the impedance seen looking into the antenna from the saw filter will vary more.
Also
Tx path will now have a saw as loss so output power will be reduced by the saw insertion loss
The benefits is that fcc spurious transmission requirements will be easier to be met.
do you have an s2p file of the saw filter and switch if so check out my simulator where you can load them in and quickly try some what if scenarios for matching return loss etc.
http://rfcooltools.com