@amaclin: if the tests for standardness on testnet are not implemented, why is the error "mandatory-script-verify-flag-failed" happening? Am I missing s.th.?

Looking at TX_OUT[0] pk_script:
01022103341B4D862E7E1C3D7FE7D0ADE7E97DFAC122760B1EF4FFB21B7A9A27DD525B6C2102EE1 C79ACEE56F0E2CAFCC1069A6E1E1E9D059FF9A60B60904E97A455D5B27F672102839D639C109F09 0104EAB7C3EEF07D6BF25952A5A88C65BC5B185F0EA2A3FAF20103AE

    01: Push next byte as data onto stack
        02 (2-of-x multisig?)
    21: OP_Data33
        03341B4D862E7E1C:3D7FE7D0ADE7E97D
        FAC122760B1EF4FF:B21B7A9A27DD525B
        6C
    21: OP_Data33
        02EE1C79ACEE56F0:E2CAFCC1069A6E1E
        1E9D059FF9A60B60:904E97A455D5B27F
        67
    21: OP_Data33
        02839D639C109F09:0104EAB7C3EEF07D
        6BF25952A5A88C65:BC5B185F0EA2A3FA
        F2
    01: Push next byte as data onto stack
        03 (2-of-3 multisig)
    AE: OP_CHECKMULTISIG

pk1=mqBRdBj2aen8aM3t8vtuSdiQLh2kSDZZfc
pk2=mwyjowbdKK9QHK6nGHTS8HoNEsULxCbNce
pk3=mqxmeXoLf3ANX6fY1epLZEHcnNvr1rYPMA

So if pgmforever assembles his tx with two priv keys of the mentioned pubkeys (the sequence of sigs must be same as the pubkeys), it should be ok? I had the impression, it fails with the mentioned error. 

@pgmforever: maybe we can see the spending tx?
It should be good to go on regtest network as well...