As others have mentioned, using two move actions to escape has just validated the grab in the first place. I would rerun the numbers with the second character using one escape chance.
Ok.
Example 1: 2 first level Fighters with 18 Str and 12 Dex, neither has Athletics or Acrobatics.
Chance to grab: 70%
Chance to escape: 45% with one attempt
Chance for this tactic to be successful for Fighter #1 for one round: 38.5%
If grabbed, chance to restrain as per his house rule: 38.5% * 55% = 21.175%
Chance to escape: 45% with one attempt
Chance for this tactic to be successful for Fighter #1 for two rounds: 11.64%
If grabbed, chance to pin as per his house rule: 11.64% * 55% = 6.4%
Chance to escape: 45% with one attempt, 69.75% with two since Fighter #2 can only try to escape at this point
Chance for this tactic to be successful for Fighter #1 for three rounds: 1.936%, less than 1 chance in 50
Example 2: 2 first level Fighters with 18 Str and 12 Dex, both have Athletics.
Chance to grab: 70%
Chance to escape: 70% with one attempt
Chance for this tactic to be successful for Fighter #1 for one round: 21%
If grabbed, chance to restrain as per his house rule: 21% * 55% = 11.55%
Chance to escape: 70% with one attempt
Chance for this tactic to be successful for Fighter #1 for two rounds: 3.465%
If grabbed, chance to pin as per his house rule: 3.465% * 55% = 1.9%
Chance to escape: 70% with one attempt, 91% with two since Fighter #2 can only try to escape at this point
Chance for this tactic to be successful for Fighter #1 for three rounds: 0.171%, less than 1 chance in 584
Example 3: 1 first level Fighter with 18 Str grabbing a first level Wizard with 10 Dex, 10 Con, and no skill:
Chance to grab: 75%
Chance to escape: 25% with one attempt
Chance for this tactic to be successful for Fighter #1 for one round: 56.25%
If grabbed, chance to restrain as per his house rule: 56.25% * 75% = 42.1875%
Chance to escape: 25% with one attempt
Chance for this tactic to be successful for Fighter #1 for two rounds: 31.64%
If grabbed, chance to pin as per his house rule: 31.64% * 75% = 23.7%
Chance to escape: 25% with one attempt, 44.75% with two since the Wizard can only try to escape at this point
Chance for this tactic to be successful for Fighter #1 for three rounds: 13.1%, less than 1 chance in 7 against a super wimpy Wizard
In all of these cases, the attacker barely has a chance to pin the target. The target has multiple chances to attack and damage the attacker. And any successful attack by the defender that moves the attacker or dazes/stuns/knocks unconscious the attacker breaks the grab/restrain/pin. This does even include situations where the target can self teleport or shift with a power.
Explain again how these are good odds for the attacker?
Note: Escape attacks are skill checks, not attack rolls, so they do not get penalized for restrained.