Is the block roll modified in any way? If not, it's going to usually be successful for combatants with shield skill, since it's opposing an attack roll which has been modified down by DB and parry.
Let's see what happens, looking on the Broadsword table (I'm looking at the RMFRP tables, 1999 printing, but I think it is close enough to the same):
Against AT 1:
* a roll of 90 does 11B. Your full shield reduction brings that down to 0. A +25 shield DB bonus would also bring it to 0.
* a roll of 110 does 18D, your full shield on a block brings that down to 0B. A +25 shield DB would bring it down to 12C.
* a roll of 130 does 24E, yours brings that down to 0C, DB 25 brings it down to 16D.
* a roll of 150 does 30E. Yours: 5C. DB: 22E.
AT 14:
* roll of 90, 7. Yours: 0. DB: 3.
* roll of 110, 11A. Yours: 0. DB: 6.
* roll of 130, 15C. Yours: 0A. DB: 10AS.
* roll of 150, 18E. Yours: 0C. DB: 14C.
AT 20:
* roll of 90, 5. Yours: 0. DB: 3.
* roll of 110, 6. Yours: 0. DB: 4.
* roll of 130, 7A. Yours: 0. DB: 6.
* roll of 150, 8E. Yours: 0C. DB: 7AK.
That's only a sampling, but in general your method blocks all or nearly all the hits, and some of the criticals (when the block roll succeeds). Against AT 14, the criticals are actually fairly similar either way. Against AT 1, the criticals are less severe. Against AT 20, the criticals on harder hits may be more severe than if the shield only gave DB -- also, the harder hits are the ones where you are least likely to block successfully, because the target number is higher. Heavy armor may be losing out here (especially since the block will sometimes fail). But light armor is coming out ahead already, so just giving better benefits for the shield may not be the best answer. An all-or-nothing block would treat all armor types the same, but then you need to adjust the odds of a successful block a little more carefully.
The other thing to think about is that because your method is cancelling most of the hits, but not the crits, fights will more often be decided by a lucky crit roll and less often by attrition. Personally I consider that not advantageous, but opinions vary and it's possible it usually goes that way for you already.