The new rules went into place in 2012. The MLB draft tracker has everyone's bonuses back through 2017, making it fairly easy to tabulate. Here is a breakdown of the Orioles spending.
2021 - $11.8293M pool - spent $11.6697M - under by $159.6k - 5% overage of $591k: all unspent
2020 - $13.8943M pool - spent $13.6800M - under by $214.3k - 5% overage of $695k: all unspent
2019 - $13.8213M pool - spent $13.6546M - under by $166.7k - 5% overage of $691k: all unspent
2018 - $8.7544M pool - spent $8.821M - OVER by $66.6k - 5% overage of $438k: $66.6 spent + $49.9k tax
2017 - $6.4370M pool - spent $5.9243M - under by $512.7k - 5% overage of $322k unspent
I think we have been giving them too much credit in terms of spending everything they can on the draft. They have not. No teams go over the 5% limit after which future draft picks are taken away, but teams spend over their pools up to that 5% mark frequently. Don't have up-to-date numbers, but one article I read prior to 2020 draft said that teams have gone over their pool 49 times (in 8 prior years under the system) including 21 of 30 teams going over in 2019. The penalty for going 0-5% over is a 75% tax on the overage. So basically if you hit your 5% over mark of $600-700k you pay an additional $450-525k to do so. All it costs is a little cash.
So the past 5 drafts the Orioles had $54.74M in bonus pool. They spent $53.80M (including a small overage tax in 2018). That makes it feel like they're spending nearly all they are allowed, but they are not. They could've given out an extra 5% in bonuses ($2.737M) by paying those bonuses plus a 5% tax ($2.05M). They're cheaping out by about a million bucks a year.
I don't know how much this matters, but being able to sign one more mid-round overslot guy like Mayo or Baumler or Willems every year seems like it could be a big deal. The Orioles are simply deciding not to sign a guy like that to add to the organization because they think the extra million it would cost is not worth it.