Yes but whether something is of poor design or not is a matter of opinion also.
A design is not objectively "good" or "poor" until everyone defines and agrees to some objective criteria by which to make the evaluation.
"Matter of opinion" is a fairly elastic phrase.
Whether chocolate is tastier than caramel as an icecream or milkshake flavouring is a matter opinion.
Whether a milkshake offers as interesting a gastronomic experience as a good quality wine is a matter of opinion.
Whether Macbeth or Hamlet is the superior play is a matter of opinion.
Whether Shakespeare or J K Rowling is the better writer is a matter of opinion.
Not all things which are matters of opinion are subject to opinion in the same way. Of the above, I think there's a very strong case to be made that Shakespeare is a better writer than J K Rowling. I also think there is a fairly strong case to be made in favour of the wine over the milkshake (which is not to say that sometimes, perhaps often, one doesn't want a milkshake rather than a wine - superior doesn't always mean preferable).
I can adduce any objective criteria by which I make the above judgments, but that doesn't shake my confidence in them.
To bring this back to RPGs, here is a design flaw in Rolemaster: The PC creation process - which is complex and involved - produces lovingly detailed, subtly differentiated characters. The PC creation process also strongly encourages (and doesn't fall very far short of guaranteeing) each PC to have some significant, combat-relevant ability. The GM guidelines in RM are fairly thin, but the monster stats are primarily relevant for combat, which together with the PC build rules strongly suggests that PCs are meant, from time to time at least, to engage in combat. And the combat resolution rules are such that low level PCs are likely often to be killed in combat, and even mid-to-high level PCs can suffer very serious consequences (although they are likely to have magic available to mitigate the worst outcomes). So if players and GMs play the game in the fashion that the various bits of the rules and guidelines suggest is the way the game is meant to be played, they will spend a good deal of their time developing
new lovingly detailed PCs who run a good chance of never actually getting to be played for very long.
This is a flaw. It's not a fatal flaw - different RM groups develop their own workarounds, and I personally have got nearly 20 years of great play out of Rolemaster. But it's a flaw nevertheless. It's a flaw that is absent from other high-death games (like low-level classic D&D and Tunnels and Trolls) because those games have quick PC gen. I think it's an open question whether it's a flaw in Classic Traveller or not - Traveller does have time-consuming and lovingly detailed PC gen, and it does encourage combat as a significant ingame activity, and combat in Traveller is very deadly - but Traveller does have robust action resolution mechanics for activity other than fighting (like voyaging and trading, for example) which mark a noticeable distinction from Rolemaster.
Here's a design flaw in 3E: The character build and action resolution mechanics for skills are, on the whole, towards the RM/RQ/Traveller style "gritty" end of things. Not every PC can be good at everything, or even that many things; DCs are set in a somewhat realistic fashion (as discussed by Justin Alexander
here); except for some of the extreme Balance and Tumbling DCs in the ELH, skills are limited to what is more-or-less humanly possible. The character build and action resolution mechanics for combat, on the other hand, are anything but gritty - even 5th level PCs (who, according to Justin Alexander, at the upper end of real world human possibility) can tackle lions bare-handed, wrestle ogres and the like, and it only gets more wahoo as levels are gained. And the hit point mechanic just exaggerates this wahoo-ness.
This mix of gonzo and gritty is by no means a fatal flaw. But it tends to produce oddities in play, and makes some aspects of world design, and also challenge design, harder than they need be, because when gonzo and gritty touch one another it's not always easy to maintain coherence.
I think that 4e tries to avoid the flaws I've identified in RM and 3E by going gonzo all the way, with both skills and combat, thereby ensuring that the lovingly-built and complex PCs get plenty of screen time to do whatever it is that they're going to do. That some dislike this play experience means that 4e is not the game for them, but I don't think it means therefore that 4e itself is flawed in being gonzo.
Which is not to say that 4e doesn't have design flaws. The complexity and clunkiness of the hiding/percpetion/invisibility/concealment rules is one such. The fact that monster XP doubles every 4 levels while item gp value multiplies by 5 every 5 levels is arguably another, as it makes static encounter design impossible, which is arguably an unnecessary burden on GM prep time - although the rapidly escalating treasure value is meant to serve another purpose, of quickly making irrelevant past GM errors in over- or under-awarding treasure. And the fact that combat bonsues and DCs scale at a different rate from skill bonuses and DCs (not to mention the other much-discussed maths debates around Expertise/Improved Defences) is yet another - as it makes integrating combat and non-combat challenges that much harder.
I can't articulate any objective criteria by which I'm distinguishing my identification of design flaws from mere questions of taste, but I nevertheless think it can be done, despite nevertheless remaining (in some sense) a matter of opinion.